timm
/

Image Classification
timm
PyTorch
Safetensors
rwightman HF staff commited on
Commit
7a7773c
1 Parent(s): 309dcfe

Update model config and README

Browse files
Files changed (2) hide show
  1. README.md +75 -120
  2. model.safetensors +3 -0
README.md CHANGED
@@ -32,8 +32,9 @@ from urllib.request import urlopen
32
  from PIL import Image
33
  import timm
34
 
35
- img = Image.open(
36
- urlopen('https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'))
 
37
 
38
  model = timm.create_model('convnextv2_huge.fcmae_ft_in22k_in1k_384', pretrained=True)
39
  model = model.eval()
@@ -53,8 +54,9 @@ from urllib.request import urlopen
53
  from PIL import Image
54
  import timm
55
 
56
- img = Image.open(
57
- urlopen('https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'))
 
58
 
59
  model = timm.create_model(
60
  'convnextv2_huge.fcmae_ft_in22k_in1k_384',
@@ -71,11 +73,12 @@ output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batc
71
 
72
  for o in output:
73
  # print shape of each feature map in output
74
- # e.g. for convnext_base:
75
- # torch.Size([1, 128, 56, 56])
76
- # torch.Size([1, 256, 28, 28])
77
- # torch.Size([1, 512, 14, 14])
78
- # torch.Size([1, 1024, 7, 7])
 
79
  print(o.shape)
80
  ```
81
 
@@ -85,8 +88,9 @@ from urllib.request import urlopen
85
  from PIL import Image
86
  import timm
87
 
88
- img = Image.open(
89
- urlopen('https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'))
 
90
 
91
  model = timm.create_model(
92
  'convnextv2_huge.fcmae_ft_in22k_in1k_384',
@@ -104,122 +108,73 @@ output = model(transforms(img).unsqueeze(0)) # output is (batch_size, num_featu
104
  # or equivalently (without needing to set num_classes=0)
105
 
106
  output = model.forward_features(transforms(img).unsqueeze(0))
107
- # output is unpooled (ie.e a (batch_size, num_features, H, W) tensor
108
 
109
  output = model.forward_head(output, pre_logits=True)
110
- # output is (batch_size, num_features) tensor
111
  ```
112
 
113
  ## Model Comparison
114
- ### By Top-1
115
- All timing numbers from eager model PyTorch 1.13 on RTX 3090 w/ AMP.
116
 
117
- |model |top1 |top5 |img_size|param_count|gmacs |macts |samples_per_sec|batch_size|
118
- |----------------------------------------------|------|------|--------|-----------|------|------|---------------|----------|
119
- |[convnextv2_huge.fcmae_ft_in22k_in1k_512](https://huggingface.co/timm/convnextv2_huge.fcmae_ft_in22k_in1k_512)|88.848|98.742|512 |660.29 |600.81|413.07|28.58 |48 |
120
- |[convnextv2_huge.fcmae_ft_in22k_in1k_384](https://huggingface.co/timm/convnextv2_huge.fcmae_ft_in22k_in1k_384)|88.668|98.738|384 |660.29 |337.96|232.35|50.56 |64 |
121
- |[convnextv2_large.fcmae_ft_in22k_in1k_384](https://huggingface.co/timm/convnextv2_large.fcmae_ft_in22k_in1k_384)|88.196|98.532|384 |197.96 |101.1 |126.74|128.94 |128 |
122
- |[convnext_xlarge.fb_in22k_ft_in1k_384](https://huggingface.co/timm/convnext_xlarge.fb_in22k_ft_in1k_384)|87.75 |98.556|384 |350.2 |179.2 |168.99|124.85 |192 |
123
- |[convnextv2_base.fcmae_ft_in22k_in1k_384](https://huggingface.co/timm/convnextv2_base.fcmae_ft_in22k_in1k_384)|87.646|98.422|384 |88.72 |45.21 |84.49 |209.51 |256 |
124
- |[convnext_large.fb_in22k_ft_in1k_384](https://huggingface.co/timm/convnext_large.fb_in22k_ft_in1k_384)|87.476|98.382|384 |197.77 |101.1 |126.74|194.66 |256 |
125
- |[convnext_large_mlp.clip_laion2b_augreg_ft_in1k](https://huggingface.co/timm/convnext_large_mlp.clip_laion2b_augreg_ft_in1k)|87.344|98.218|256 |200.13 |44.94 |56.33 |438.08 |256 |
126
- |[convnextv2_large.fcmae_ft_in22k_in1k](https://huggingface.co/timm/convnextv2_large.fcmae_ft_in22k_in1k)|87.26 |98.248|224 |197.96 |34.4 |43.13 |376.84 |256 |
127
- |[convnext_xlarge.fb_in22k_ft_in1k](https://huggingface.co/timm/convnext_xlarge.fb_in22k_ft_in1k)|87.002|98.208|224 |350.2 |60.98 |57.5 |368.01 |256 |
128
- |[convnext_base.fb_in22k_ft_in1k_384](https://huggingface.co/timm/convnext_base.fb_in22k_ft_in1k_384)|86.796|98.264|384 |88.59 |45.21 |84.49 |366.54 |256 |
129
- |[convnextv2_base.fcmae_ft_in22k_in1k](https://huggingface.co/timm/convnextv2_base.fcmae_ft_in22k_in1k)|86.74 |98.022|224 |88.72 |15.38 |28.75 |624.23 |256 |
130
- |[convnext_large.fb_in22k_ft_in1k](https://huggingface.co/timm/convnext_large.fb_in22k_ft_in1k)|86.636|98.028|224 |197.77 |34.4 |43.13 |581.43 |256 |
131
- |[convnext_base.clip_laiona_augreg_ft_in1k_384](https://huggingface.co/timm/convnext_base.clip_laiona_augreg_ft_in1k_384)|86.504|97.97 |384 |88.59 |45.21 |84.49 |368.14 |256 |
132
- |[convnextv2_huge.fcmae_ft_in1k](https://huggingface.co/timm/convnextv2_huge.fcmae_ft_in1k)|86.256|97.75 |224 |660.29 |115.0 |79.07 |154.72 |256 |
133
- |[convnext_small.in12k_ft_in1k_384](https://huggingface.co/timm/convnext_small.in12k_ft_in1k_384)|86.182|97.92 |384 |50.22 |25.58 |63.37 |516.19 |256 |
134
- |[convnext_base.clip_laion2b_augreg_ft_in1k](https://huggingface.co/timm/convnext_base.clip_laion2b_augreg_ft_in1k)|86.154|97.68 |256 |88.59 |20.09 |37.55 |819.86 |256 |
135
- |[convnext_base.fb_in22k_ft_in1k](https://huggingface.co/timm/convnext_base.fb_in22k_ft_in1k)|85.822|97.866|224 |88.59 |15.38 |28.75 |1037.66 |256 |
136
- |[convnext_small.fb_in22k_ft_in1k_384](https://huggingface.co/timm/convnext_small.fb_in22k_ft_in1k_384)|85.778|97.886|384 |50.22 |25.58 |63.37 |518.95 |256 |
137
- |[convnextv2_large.fcmae_ft_in1k](https://huggingface.co/timm/convnextv2_large.fcmae_ft_in1k)|85.742|97.584|224 |197.96 |34.4 |43.13 |375.23 |256 |
138
- |[convnext_small.in12k_ft_in1k](https://huggingface.co/timm/convnext_small.in12k_ft_in1k)|85.174|97.506|224 |50.22 |8.71 |21.56 |1474.31 |256 |
139
- |[convnext_tiny.in12k_ft_in1k_384](https://huggingface.co/timm/convnext_tiny.in12k_ft_in1k_384)|85.118|97.608|384 |28.59 |13.14 |39.48 |856.76 |256 |
140
- |[convnextv2_tiny.fcmae_ft_in22k_in1k_384](https://huggingface.co/timm/convnextv2_tiny.fcmae_ft_in22k_in1k_384)|85.112|97.63 |384 |28.64 |13.14 |39.48 |491.32 |256 |
141
- |[convnextv2_base.fcmae_ft_in1k](https://huggingface.co/timm/convnextv2_base.fcmae_ft_in1k)|84.874|97.09 |224 |88.72 |15.38 |28.75 |625.33 |256 |
142
- |[convnext_small.fb_in22k_ft_in1k](https://huggingface.co/timm/convnext_small.fb_in22k_ft_in1k)|84.562|97.394|224 |50.22 |8.71 |21.56 |1478.29 |256 |
143
- |[convnext_large.fb_in1k](https://huggingface.co/timm/convnext_large.fb_in1k)|84.282|96.892|224 |197.77 |34.4 |43.13 |584.28 |256 |
144
- |[convnext_tiny.in12k_ft_in1k](https://huggingface.co/timm/convnext_tiny.in12k_ft_in1k)|84.186|97.124|224 |28.59 |4.47 |13.44 |2433.7 |256 |
145
- |[convnext_tiny.fb_in22k_ft_in1k_384](https://huggingface.co/timm/convnext_tiny.fb_in22k_ft_in1k_384)|84.084|97.14 |384 |28.59 |13.14 |39.48 |862.95 |256 |
146
- |[convnextv2_tiny.fcmae_ft_in22k_in1k](https://huggingface.co/timm/convnextv2_tiny.fcmae_ft_in22k_in1k)|83.894|96.964|224 |28.64 |4.47 |13.44 |1452.72 |256 |
147
- |[convnext_base.fb_in1k](https://huggingface.co/timm/convnext_base.fb_in1k)|83.82 |96.746|224 |88.59 |15.38 |28.75 |1054.0 |256 |
148
- |[convnextv2_nano.fcmae_ft_in22k_in1k_384](https://huggingface.co/timm/convnextv2_nano.fcmae_ft_in22k_in1k_384)|83.37 |96.742|384 |15.62 |7.22 |24.61 |801.72 |256 |
149
- |[convnext_small.fb_in1k](https://huggingface.co/timm/convnext_small.fb_in1k)|83.142|96.434|224 |50.22 |8.71 |21.56 |1464.0 |256 |
150
- |[convnextv2_tiny.fcmae_ft_in1k](https://huggingface.co/timm/convnextv2_tiny.fcmae_ft_in1k)|82.92 |96.284|224 |28.64 |4.47 |13.44 |1425.62 |256 |
151
- |[convnext_tiny.fb_in22k_ft_in1k](https://huggingface.co/timm/convnext_tiny.fb_in22k_ft_in1k)|82.898|96.616|224 |28.59 |4.47 |13.44 |2480.88 |256 |
152
- |[convnext_nano.in12k_ft_in1k](https://huggingface.co/timm/convnext_nano.in12k_ft_in1k)|82.282|96.344|224 |15.59 |2.46 |8.37 |3926.52 |256 |
153
- |[convnext_tiny_hnf.a2h_in1k](https://huggingface.co/timm/convnext_tiny_hnf.a2h_in1k)|82.216|95.852|224 |28.59 |4.47 |13.44 |2529.75 |256 |
154
- |[convnext_tiny.fb_in1k](https://huggingface.co/timm/convnext_tiny.fb_in1k)|82.066|95.854|224 |28.59 |4.47 |13.44 |2346.26 |256 |
155
- |[convnextv2_nano.fcmae_ft_in22k_in1k](https://huggingface.co/timm/convnextv2_nano.fcmae_ft_in22k_in1k)|82.03 |96.166|224 |15.62 |2.46 |8.37 |2300.18 |256 |
156
- |[convnextv2_nano.fcmae_ft_in1k](https://huggingface.co/timm/convnextv2_nano.fcmae_ft_in1k)|81.83 |95.738|224 |15.62 |2.46 |8.37 |2321.48 |256 |
157
- |[convnext_nano_ols.d1h_in1k](https://huggingface.co/timm/convnext_nano_ols.d1h_in1k)|80.866|95.246|224 |15.65 |2.65 |9.38 |3523.85 |256 |
158
- |[convnext_nano.d1h_in1k](https://huggingface.co/timm/convnext_nano.d1h_in1k)|80.768|95.334|224 |15.59 |2.46 |8.37 |3915.58 |256 |
159
- |[convnextv2_pico.fcmae_ft_in1k](https://huggingface.co/timm/convnextv2_pico.fcmae_ft_in1k)|80.304|95.072|224 |9.07 |1.37 |6.1 |3274.57 |256 |
160
- |[convnext_pico.d1_in1k](https://huggingface.co/timm/convnext_pico.d1_in1k)|79.526|94.558|224 |9.05 |1.37 |6.1 |5686.88 |256 |
161
- |[convnext_pico_ols.d1_in1k](https://huggingface.co/timm/convnext_pico_ols.d1_in1k)|79.522|94.692|224 |9.06 |1.43 |6.5 |5422.46 |256 |
162
- |[convnextv2_femto.fcmae_ft_in1k](https://huggingface.co/timm/convnextv2_femto.fcmae_ft_in1k)|78.488|93.98 |224 |5.23 |0.79 |4.57 |4264.2 |256 |
163
- |[convnext_femto_ols.d1_in1k](https://huggingface.co/timm/convnext_femto_ols.d1_in1k)|77.86 |93.83 |224 |5.23 |0.82 |4.87 |6910.6 |256 |
164
- |[convnext_femto.d1_in1k](https://huggingface.co/timm/convnext_femto.d1_in1k)|77.454|93.68 |224 |5.22 |0.79 |4.57 |7189.92 |256 |
165
- |[convnextv2_atto.fcmae_ft_in1k](https://huggingface.co/timm/convnextv2_atto.fcmae_ft_in1k)|76.664|93.044|224 |3.71 |0.55 |3.81 |4728.91 |256 |
166
- |[convnext_atto_ols.a2_in1k](https://huggingface.co/timm/convnext_atto_ols.a2_in1k)|75.88 |92.846|224 |3.7 |0.58 |4.11 |7963.16 |256 |
167
- |[convnext_atto.d2_in1k](https://huggingface.co/timm/convnext_atto.d2_in1k)|75.664|92.9 |224 |3.7 |0.55 |3.81 |8439.22 |256 |
168
-
169
- ### By Throughput (samples / sec)
170
  All timing numbers from eager model PyTorch 1.13 on RTX 3090 w/ AMP.
171
 
172
- |model |top1 |top5 |img_size|param_count|gmacs |macts |samples_per_sec|batch_size|
173
- |----------------------------------------------|------|------|--------|-----------|------|------|---------------|----------|
174
- |[convnext_atto.d2_in1k](https://huggingface.co/timm/convnext_atto.d2_in1k)|75.664|92.9 |224 |3.7 |0.55 |3.81 |8439.22 |256 |
175
- |[convnext_atto_ols.a2_in1k](https://huggingface.co/timm/convnext_atto_ols.a2_in1k)|75.88 |92.846|224 |3.7 |0.58 |4.11 |7963.16 |256 |
176
- |[convnext_femto.d1_in1k](https://huggingface.co/timm/convnext_femto.d1_in1k)|77.454|93.68 |224 |5.22 |0.79 |4.57 |7189.92 |256 |
177
- |[convnext_femto_ols.d1_in1k](https://huggingface.co/timm/convnext_femto_ols.d1_in1k)|77.86 |93.83 |224 |5.23 |0.82 |4.87 |6910.6 |256 |
178
- |[convnext_pico.d1_in1k](https://huggingface.co/timm/convnext_pico.d1_in1k)|79.526|94.558|224 |9.05 |1.37 |6.1 |5686.88 |256 |
179
- |[convnext_pico_ols.d1_in1k](https://huggingface.co/timm/convnext_pico_ols.d1_in1k)|79.522|94.692|224 |9.06 |1.43 |6.5 |5422.46 |256 |
180
- |[convnextv2_atto.fcmae_ft_in1k](https://huggingface.co/timm/convnextv2_atto.fcmae_ft_in1k)|76.664|93.044|224 |3.71 |0.55 |3.81 |4728.91 |256 |
181
- |[convnextv2_femto.fcmae_ft_in1k](https://huggingface.co/timm/convnextv2_femto.fcmae_ft_in1k)|78.488|93.98 |224 |5.23 |0.79 |4.57 |4264.2 |256 |
182
- |[convnext_nano.in12k_ft_in1k](https://huggingface.co/timm/convnext_nano.in12k_ft_in1k)|82.282|96.344|224 |15.59 |2.46 |8.37 |3926.52 |256 |
183
- |[convnext_nano.d1h_in1k](https://huggingface.co/timm/convnext_nano.d1h_in1k)|80.768|95.334|224 |15.59 |2.46 |8.37 |3915.58 |256 |
184
- |[convnext_nano_ols.d1h_in1k](https://huggingface.co/timm/convnext_nano_ols.d1h_in1k)|80.866|95.246|224 |15.65 |2.65 |9.38 |3523.85 |256 |
185
- |[convnextv2_pico.fcmae_ft_in1k](https://huggingface.co/timm/convnextv2_pico.fcmae_ft_in1k)|80.304|95.072|224 |9.07 |1.37 |6.1 |3274.57 |256 |
186
- |[convnext_tiny_hnf.a2h_in1k](https://huggingface.co/timm/convnext_tiny_hnf.a2h_in1k)|82.216|95.852|224 |28.59 |4.47 |13.44 |2529.75 |256 |
187
- |[convnext_tiny.fb_in22k_ft_in1k](https://huggingface.co/timm/convnext_tiny.fb_in22k_ft_in1k)|82.898|96.616|224 |28.59 |4.47 |13.44 |2480.88 |256 |
188
- |[convnext_tiny.in12k_ft_in1k](https://huggingface.co/timm/convnext_tiny.in12k_ft_in1k)|84.186|97.124|224 |28.59 |4.47 |13.44 |2433.7 |256 |
189
- |[convnext_tiny.fb_in1k](https://huggingface.co/timm/convnext_tiny.fb_in1k)|82.066|95.854|224 |28.59 |4.47 |13.44 |2346.26 |256 |
190
- |[convnextv2_nano.fcmae_ft_in1k](https://huggingface.co/timm/convnextv2_nano.fcmae_ft_in1k)|81.83 |95.738|224 |15.62 |2.46 |8.37 |2321.48 |256 |
191
- |[convnextv2_nano.fcmae_ft_in22k_in1k](https://huggingface.co/timm/convnextv2_nano.fcmae_ft_in22k_in1k)|82.03 |96.166|224 |15.62 |2.46 |8.37 |2300.18 |256 |
192
- |[convnext_small.fb_in22k_ft_in1k](https://huggingface.co/timm/convnext_small.fb_in22k_ft_in1k)|84.562|97.394|224 |50.22 |8.71 |21.56 |1478.29 |256 |
193
- |[convnext_small.in12k_ft_in1k](https://huggingface.co/timm/convnext_small.in12k_ft_in1k)|85.174|97.506|224 |50.22 |8.71 |21.56 |1474.31 |256 |
194
- |[convnext_small.fb_in1k](https://huggingface.co/timm/convnext_small.fb_in1k)|83.142|96.434|224 |50.22 |8.71 |21.56 |1464.0 |256 |
195
- |[convnextv2_tiny.fcmae_ft_in22k_in1k](https://huggingface.co/timm/convnextv2_tiny.fcmae_ft_in22k_in1k)|83.894|96.964|224 |28.64 |4.47 |13.44 |1452.72 |256 |
196
- |[convnextv2_tiny.fcmae_ft_in1k](https://huggingface.co/timm/convnextv2_tiny.fcmae_ft_in1k)|82.92 |96.284|224 |28.64 |4.47 |13.44 |1425.62 |256 |
197
- |[convnext_base.fb_in1k](https://huggingface.co/timm/convnext_base.fb_in1k)|83.82 |96.746|224 |88.59 |15.38 |28.75 |1054.0 |256 |
198
- |[convnext_base.fb_in22k_ft_in1k](https://huggingface.co/timm/convnext_base.fb_in22k_ft_in1k)|85.822|97.866|224 |88.59 |15.38 |28.75 |1037.66 |256 |
199
- |[convnext_tiny.fb_in22k_ft_in1k_384](https://huggingface.co/timm/convnext_tiny.fb_in22k_ft_in1k_384)|84.084|97.14 |384 |28.59 |13.14 |39.48 |862.95 |256 |
200
- |[convnext_tiny.in12k_ft_in1k_384](https://huggingface.co/timm/convnext_tiny.in12k_ft_in1k_384)|85.118|97.608|384 |28.59 |13.14 |39.48 |856.76 |256 |
201
- |[convnext_base.clip_laion2b_augreg_ft_in1k](https://huggingface.co/timm/convnext_base.clip_laion2b_augreg_ft_in1k)|86.154|97.68 |256 |88.59 |20.09 |37.55 |819.86 |256 |
202
- |[convnextv2_nano.fcmae_ft_in22k_in1k_384](https://huggingface.co/timm/convnextv2_nano.fcmae_ft_in22k_in1k_384)|83.37 |96.742|384 |15.62 |7.22 |24.61 |801.72 |256 |
203
- |[convnextv2_base.fcmae_ft_in1k](https://huggingface.co/timm/convnextv2_base.fcmae_ft_in1k)|84.874|97.09 |224 |88.72 |15.38 |28.75 |625.33 |256 |
204
- |[convnextv2_base.fcmae_ft_in22k_in1k](https://huggingface.co/timm/convnextv2_base.fcmae_ft_in22k_in1k)|86.74 |98.022|224 |88.72 |15.38 |28.75 |624.23 |256 |
205
- |[convnext_large.fb_in1k](https://huggingface.co/timm/convnext_large.fb_in1k)|84.282|96.892|224 |197.77 |34.4 |43.13 |584.28 |256 |
206
- |[convnext_large.fb_in22k_ft_in1k](https://huggingface.co/timm/convnext_large.fb_in22k_ft_in1k)|86.636|98.028|224 |197.77 |34.4 |43.13 |581.43 |256 |
207
- |[convnext_small.fb_in22k_ft_in1k_384](https://huggingface.co/timm/convnext_small.fb_in22k_ft_in1k_384)|85.778|97.886|384 |50.22 |25.58 |63.37 |518.95 |256 |
208
- |[convnext_small.in12k_ft_in1k_384](https://huggingface.co/timm/convnext_small.in12k_ft_in1k_384)|86.182|97.92 |384 |50.22 |25.58 |63.37 |516.19 |256 |
209
- |[convnextv2_tiny.fcmae_ft_in22k_in1k_384](https://huggingface.co/timm/convnextv2_tiny.fcmae_ft_in22k_in1k_384)|85.112|97.63 |384 |28.64 |13.14 |39.48 |491.32 |256 |
210
- |[convnext_large_mlp.clip_laion2b_augreg_ft_in1k](https://huggingface.co/timm/convnext_large_mlp.clip_laion2b_augreg_ft_in1k)|87.344|98.218|256 |200.13 |44.94 |56.33 |438.08 |256 |
211
- |[convnextv2_large.fcmae_ft_in22k_in1k](https://huggingface.co/timm/convnextv2_large.fcmae_ft_in22k_in1k)|87.26 |98.248|224 |197.96 |34.4 |43.13 |376.84 |256 |
212
- |[convnextv2_large.fcmae_ft_in1k](https://huggingface.co/timm/convnextv2_large.fcmae_ft_in1k)|85.742|97.584|224 |197.96 |34.4 |43.13 |375.23 |256 |
213
- |[convnext_base.clip_laiona_augreg_ft_in1k_384](https://huggingface.co/timm/convnext_base.clip_laiona_augreg_ft_in1k_384)|86.504|97.97 |384 |88.59 |45.21 |84.49 |368.14 |256 |
214
- |[convnext_xlarge.fb_in22k_ft_in1k](https://huggingface.co/timm/convnext_xlarge.fb_in22k_ft_in1k)|87.002|98.208|224 |350.2 |60.98 |57.5 |368.01 |256 |
215
- |[convnext_base.fb_in22k_ft_in1k_384](https://huggingface.co/timm/convnext_base.fb_in22k_ft_in1k_384)|86.796|98.264|384 |88.59 |45.21 |84.49 |366.54 |256 |
216
- |[convnextv2_base.fcmae_ft_in22k_in1k_384](https://huggingface.co/timm/convnextv2_base.fcmae_ft_in22k_in1k_384)|87.646|98.422|384 |88.72 |45.21 |84.49 |209.51 |256 |
217
- |[convnext_large.fb_in22k_ft_in1k_384](https://huggingface.co/timm/convnext_large.fb_in22k_ft_in1k_384)|87.476|98.382|384 |197.77 |101.1 |126.74|194.66 |256 |
218
- |[convnextv2_huge.fcmae_ft_in1k](https://huggingface.co/timm/convnextv2_huge.fcmae_ft_in1k)|86.256|97.75 |224 |660.29 |115.0 |79.07 |154.72 |256 |
219
- |[convnextv2_large.fcmae_ft_in22k_in1k_384](https://huggingface.co/timm/convnextv2_large.fcmae_ft_in22k_in1k_384)|88.196|98.532|384 |197.96 |101.1 |126.74|128.94 |128 |
220
- |[convnext_xlarge.fb_in22k_ft_in1k_384](https://huggingface.co/timm/convnext_xlarge.fb_in22k_ft_in1k_384)|87.75 |98.556|384 |350.2 |179.2 |168.99|124.85 |192 |
221
- |[convnextv2_huge.fcmae_ft_in22k_in1k_384](https://huggingface.co/timm/convnextv2_huge.fcmae_ft_in22k_in1k_384)|88.668|98.738|384 |660.29 |337.96|232.35|50.56 |64 |
222
- |[convnextv2_huge.fcmae_ft_in22k_in1k_512](https://huggingface.co/timm/convnextv2_huge.fcmae_ft_in22k_in1k_512)|88.848|98.742|512 |660.29 |600.81|413.07|28.58 |48 |
 
 
 
 
 
223
 
224
  ## Citation
225
  ```bibtex
@@ -238,6 +193,6 @@ All timing numbers from eager model PyTorch 1.13 on RTX 3090 w/ AMP.
238
  publisher = {GitHub},
239
  journal = {GitHub repository},
240
  doi = {10.5281/zenodo.4414861},
241
- howpublished = {\url{https://github.com/rwightman/pytorch-image-models}}
242
  }
243
  ```
 
32
  from PIL import Image
33
  import timm
34
 
35
+ img = Image.open(urlopen(
36
+ 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'
37
+ ))
38
 
39
  model = timm.create_model('convnextv2_huge.fcmae_ft_in22k_in1k_384', pretrained=True)
40
  model = model.eval()
 
54
  from PIL import Image
55
  import timm
56
 
57
+ img = Image.open(urlopen(
58
+ 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'
59
+ ))
60
 
61
  model = timm.create_model(
62
  'convnextv2_huge.fcmae_ft_in22k_in1k_384',
 
73
 
74
  for o in output:
75
  # print shape of each feature map in output
76
+ # e.g.:
77
+ # torch.Size([1, 352, 96, 96])
78
+ # torch.Size([1, 704, 48, 48])
79
+ # torch.Size([1, 1408, 24, 24])
80
+ # torch.Size([1, 2816, 12, 12])
81
+
82
  print(o.shape)
83
  ```
84
 
 
88
  from PIL import Image
89
  import timm
90
 
91
+ img = Image.open(urlopen(
92
+ 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'
93
+ ))
94
 
95
  model = timm.create_model(
96
  'convnextv2_huge.fcmae_ft_in22k_in1k_384',
 
108
  # or equivalently (without needing to set num_classes=0)
109
 
110
  output = model.forward_features(transforms(img).unsqueeze(0))
111
+ # output is unpooled, a (1, 2816, 12, 12) shaped tensor
112
 
113
  output = model.forward_head(output, pre_logits=True)
114
+ # output is a (1, num_features) shaped tensor
115
  ```
116
 
117
  ## Model Comparison
118
+ Explore the dataset and runtime metrics of this model in timm [model results](https://github.com/huggingface/pytorch-image-models/tree/main/results).
 
119
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
120
  All timing numbers from eager model PyTorch 1.13 on RTX 3090 w/ AMP.
121
 
122
+ | model |top1 |top5 |img_size|param_count|gmacs |macts |samples_per_sec|batch_size|
123
+ |------------------------------------------------------------------------------------------------------------------------------|------|------|--------|-----------|------|------|---------------|----------|
124
+ | [convnextv2_huge.fcmae_ft_in22k_in1k_512](https://huggingface.co/timm/convnextv2_huge.fcmae_ft_in22k_in1k_512) |88.848|98.742|512 |660.29 |600.81|413.07|28.58 |48 |
125
+ | [convnextv2_huge.fcmae_ft_in22k_in1k_384](https://huggingface.co/timm/convnextv2_huge.fcmae_ft_in22k_in1k_384) |88.668|98.738|384 |660.29 |337.96|232.35|50.56 |64 |
126
+ | [convnext_xxlarge.clip_laion2b_soup_ft_in1k](https://huggingface.co/timm/convnext_xxlarge.clip_laion2b_soup_ft_in1k) |88.612|98.704|256 |846.47 |198.09|124.45|122.45 |256 |
127
+ | [convnext_large_mlp.clip_laion2b_soup_ft_in12k_in1k_384](https://huggingface.co/timm/convnext_large_mlp.clip_laion2b_soup_ft_in12k_in1k_384) |88.312|98.578|384 |200.13 |101.11|126.74|196.84 |256 |
128
+ | [convnextv2_large.fcmae_ft_in22k_in1k_384](https://huggingface.co/timm/convnextv2_large.fcmae_ft_in22k_in1k_384) |88.196|98.532|384 |197.96 |101.1 |126.74|128.94 |128 |
129
+ | [convnext_large_mlp.clip_laion2b_soup_ft_in12k_in1k_320](https://huggingface.co/timm/convnext_large_mlp.clip_laion2b_soup_ft_in12k_in1k_320) |87.968|98.47 |320 |200.13 |70.21 |88.02 |283.42 |256 |
130
+ | [convnext_xlarge.fb_in22k_ft_in1k_384](https://huggingface.co/timm/convnext_xlarge.fb_in22k_ft_in1k_384) |87.75 |98.556|384 |350.2 |179.2 |168.99|124.85 |192 |
131
+ | [convnextv2_base.fcmae_ft_in22k_in1k_384](https://huggingface.co/timm/convnextv2_base.fcmae_ft_in22k_in1k_384) |87.646|98.422|384 |88.72 |45.21 |84.49 |209.51 |256 |
132
+ | [convnext_large.fb_in22k_ft_in1k_384](https://huggingface.co/timm/convnext_large.fb_in22k_ft_in1k_384) |87.476|98.382|384 |197.77 |101.1 |126.74|194.66 |256 |
133
+ | [convnext_large_mlp.clip_laion2b_augreg_ft_in1k](https://huggingface.co/timm/convnext_large_mlp.clip_laion2b_augreg_ft_in1k) |87.344|98.218|256 |200.13 |44.94 |56.33 |438.08 |256 |
134
+ | [convnextv2_large.fcmae_ft_in22k_in1k](https://huggingface.co/timm/convnextv2_large.fcmae_ft_in22k_in1k) |87.26 |98.248|224 |197.96 |34.4 |43.13 |376.84 |256 |
135
+ | [convnext_base.clip_laion2b_augreg_ft_in12k_in1k_384](https://huggingface.co/timm/convnext_base.clip_laion2b_augreg_ft_in12k_in1k_384) |87.138|98.212|384 |88.59 |45.21 |84.49 |365.47 |256 |
136
+ | [convnext_xlarge.fb_in22k_ft_in1k](https://huggingface.co/timm/convnext_xlarge.fb_in22k_ft_in1k) |87.002|98.208|224 |350.2 |60.98 |57.5 |368.01 |256 |
137
+ | [convnext_base.fb_in22k_ft_in1k_384](https://huggingface.co/timm/convnext_base.fb_in22k_ft_in1k_384) |86.796|98.264|384 |88.59 |45.21 |84.49 |366.54 |256 |
138
+ | [convnextv2_base.fcmae_ft_in22k_in1k](https://huggingface.co/timm/convnextv2_base.fcmae_ft_in22k_in1k) |86.74 |98.022|224 |88.72 |15.38 |28.75 |624.23 |256 |
139
+ | [convnext_large.fb_in22k_ft_in1k](https://huggingface.co/timm/convnext_large.fb_in22k_ft_in1k) |86.636|98.028|224 |197.77 |34.4 |43.13 |581.43 |256 |
140
+ | [convnext_base.clip_laiona_augreg_ft_in1k_384](https://huggingface.co/timm/convnext_base.clip_laiona_augreg_ft_in1k_384) |86.504|97.97 |384 |88.59 |45.21 |84.49 |368.14 |256 |
141
+ | [convnext_base.clip_laion2b_augreg_ft_in12k_in1k](https://huggingface.co/timm/convnext_base.clip_laion2b_augreg_ft_in12k_in1k) |86.344|97.97 |256 |88.59 |20.09 |37.55 |816.14 |256 |
142
+ | [convnextv2_huge.fcmae_ft_in1k](https://huggingface.co/timm/convnextv2_huge.fcmae_ft_in1k) |86.256|97.75 |224 |660.29 |115.0 |79.07 |154.72 |256 |
143
+ | [convnext_small.in12k_ft_in1k_384](https://huggingface.co/timm/convnext_small.in12k_ft_in1k_384) |86.182|97.92 |384 |50.22 |25.58 |63.37 |516.19 |256 |
144
+ | [convnext_base.clip_laion2b_augreg_ft_in1k](https://huggingface.co/timm/convnext_base.clip_laion2b_augreg_ft_in1k) |86.154|97.68 |256 |88.59 |20.09 |37.55 |819.86 |256 |
145
+ | [convnext_base.fb_in22k_ft_in1k](https://huggingface.co/timm/convnext_base.fb_in22k_ft_in1k) |85.822|97.866|224 |88.59 |15.38 |28.75 |1037.66 |256 |
146
+ | [convnext_small.fb_in22k_ft_in1k_384](https://huggingface.co/timm/convnext_small.fb_in22k_ft_in1k_384) |85.778|97.886|384 |50.22 |25.58 |63.37 |518.95 |256 |
147
+ | [convnextv2_large.fcmae_ft_in1k](https://huggingface.co/timm/convnextv2_large.fcmae_ft_in1k) |85.742|97.584|224 |197.96 |34.4 |43.13 |375.23 |256 |
148
+ | [convnext_small.in12k_ft_in1k](https://huggingface.co/timm/convnext_small.in12k_ft_in1k) |85.174|97.506|224 |50.22 |8.71 |21.56 |1474.31 |256 |
149
+ | [convnext_tiny.in12k_ft_in1k_384](https://huggingface.co/timm/convnext_tiny.in12k_ft_in1k_384) |85.118|97.608|384 |28.59 |13.14 |39.48 |856.76 |256 |
150
+ | [convnextv2_tiny.fcmae_ft_in22k_in1k_384](https://huggingface.co/timm/convnextv2_tiny.fcmae_ft_in22k_in1k_384) |85.112|97.63 |384 |28.64 |13.14 |39.48 |491.32 |256 |
151
+ | [convnextv2_base.fcmae_ft_in1k](https://huggingface.co/timm/convnextv2_base.fcmae_ft_in1k) |84.874|97.09 |224 |88.72 |15.38 |28.75 |625.33 |256 |
152
+ | [convnext_small.fb_in22k_ft_in1k](https://huggingface.co/timm/convnext_small.fb_in22k_ft_in1k) |84.562|97.394|224 |50.22 |8.71 |21.56 |1478.29 |256 |
153
+ | [convnext_large.fb_in1k](https://huggingface.co/timm/convnext_large.fb_in1k) |84.282|96.892|224 |197.77 |34.4 |43.13 |584.28 |256 |
154
+ | [convnext_tiny.in12k_ft_in1k](https://huggingface.co/timm/convnext_tiny.in12k_ft_in1k) |84.186|97.124|224 |28.59 |4.47 |13.44 |2433.7 |256 |
155
+ | [convnext_tiny.fb_in22k_ft_in1k_384](https://huggingface.co/timm/convnext_tiny.fb_in22k_ft_in1k_384) |84.084|97.14 |384 |28.59 |13.14 |39.48 |862.95 |256 |
156
+ | [convnextv2_tiny.fcmae_ft_in22k_in1k](https://huggingface.co/timm/convnextv2_tiny.fcmae_ft_in22k_in1k) |83.894|96.964|224 |28.64 |4.47 |13.44 |1452.72 |256 |
157
+ | [convnext_base.fb_in1k](https://huggingface.co/timm/convnext_base.fb_in1k) |83.82 |96.746|224 |88.59 |15.38 |28.75 |1054.0 |256 |
158
+ | [convnextv2_nano.fcmae_ft_in22k_in1k_384](https://huggingface.co/timm/convnextv2_nano.fcmae_ft_in22k_in1k_384) |83.37 |96.742|384 |15.62 |7.22 |24.61 |801.72 |256 |
159
+ | [convnext_small.fb_in1k](https://huggingface.co/timm/convnext_small.fb_in1k) |83.142|96.434|224 |50.22 |8.71 |21.56 |1464.0 |256 |
160
+ | [convnextv2_tiny.fcmae_ft_in1k](https://huggingface.co/timm/convnextv2_tiny.fcmae_ft_in1k) |82.92 |96.284|224 |28.64 |4.47 |13.44 |1425.62 |256 |
161
+ | [convnext_tiny.fb_in22k_ft_in1k](https://huggingface.co/timm/convnext_tiny.fb_in22k_ft_in1k) |82.898|96.616|224 |28.59 |4.47 |13.44 |2480.88 |256 |
162
+ | [convnext_nano.in12k_ft_in1k](https://huggingface.co/timm/convnext_nano.in12k_ft_in1k) |82.282|96.344|224 |15.59 |2.46 |8.37 |3926.52 |256 |
163
+ | [convnext_tiny_hnf.a2h_in1k](https://huggingface.co/timm/convnext_tiny_hnf.a2h_in1k) |82.216|95.852|224 |28.59 |4.47 |13.44 |2529.75 |256 |
164
+ | [convnext_tiny.fb_in1k](https://huggingface.co/timm/convnext_tiny.fb_in1k) |82.066|95.854|224 |28.59 |4.47 |13.44 |2346.26 |256 |
165
+ | [convnextv2_nano.fcmae_ft_in22k_in1k](https://huggingface.co/timm/convnextv2_nano.fcmae_ft_in22k_in1k) |82.03 |96.166|224 |15.62 |2.46 |8.37 |2300.18 |256 |
166
+ | [convnextv2_nano.fcmae_ft_in1k](https://huggingface.co/timm/convnextv2_nano.fcmae_ft_in1k) |81.83 |95.738|224 |15.62 |2.46 |8.37 |2321.48 |256 |
167
+ | [convnext_nano_ols.d1h_in1k](https://huggingface.co/timm/convnext_nano_ols.d1h_in1k) |80.866|95.246|224 |15.65 |2.65 |9.38 |3523.85 |256 |
168
+ | [convnext_nano.d1h_in1k](https://huggingface.co/timm/convnext_nano.d1h_in1k) |80.768|95.334|224 |15.59 |2.46 |8.37 |3915.58 |256 |
169
+ | [convnextv2_pico.fcmae_ft_in1k](https://huggingface.co/timm/convnextv2_pico.fcmae_ft_in1k) |80.304|95.072|224 |9.07 |1.37 |6.1 |3274.57 |256 |
170
+ | [convnext_pico.d1_in1k](https://huggingface.co/timm/convnext_pico.d1_in1k) |79.526|94.558|224 |9.05 |1.37 |6.1 |5686.88 |256 |
171
+ | [convnext_pico_ols.d1_in1k](https://huggingface.co/timm/convnext_pico_ols.d1_in1k) |79.522|94.692|224 |9.06 |1.43 |6.5 |5422.46 |256 |
172
+ | [convnextv2_femto.fcmae_ft_in1k](https://huggingface.co/timm/convnextv2_femto.fcmae_ft_in1k) |78.488|93.98 |224 |5.23 |0.79 |4.57 |4264.2 |256 |
173
+ | [convnext_femto_ols.d1_in1k](https://huggingface.co/timm/convnext_femto_ols.d1_in1k) |77.86 |93.83 |224 |5.23 |0.82 |4.87 |6910.6 |256 |
174
+ | [convnext_femto.d1_in1k](https://huggingface.co/timm/convnext_femto.d1_in1k) |77.454|93.68 |224 |5.22 |0.79 |4.57 |7189.92 |256 |
175
+ | [convnextv2_atto.fcmae_ft_in1k](https://huggingface.co/timm/convnextv2_atto.fcmae_ft_in1k) |76.664|93.044|224 |3.71 |0.55 |3.81 |4728.91 |256 |
176
+ | [convnext_atto_ols.a2_in1k](https://huggingface.co/timm/convnext_atto_ols.a2_in1k) |75.88 |92.846|224 |3.7 |0.58 |4.11 |7963.16 |256 |
177
+ | [convnext_atto.d2_in1k](https://huggingface.co/timm/convnext_atto.d2_in1k) |75.664|92.9 |224 |3.7 |0.55 |3.81 |8439.22 |256 |
178
 
179
  ## Citation
180
  ```bibtex
 
193
  publisher = {GitHub},
194
  journal = {GitHub repository},
195
  doi = {10.5281/zenodo.4414861},
196
+ howpublished = {\url{https://github.com/huggingface/pytorch-image-models}}
197
  }
198
  ```
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7bc97d1951c90a19f5f28fd27235e0f4536c2b82ea79e0d1afe01c6e0a0b904d
3
+ size 2641197850