timm
/

Image Classification
timm
PyTorch
Safetensors
rwightman HF staff commited on
Commit
6c3de13
1 Parent(s): 412d784

Update model config and README

Browse files
Files changed (2) hide show
  1. README.md +76 -120
  2. model.safetensors +3 -0
README.md CHANGED
@@ -13,6 +13,7 @@ datasets:
13
  A ConvNeXt image classification model. CLIP image tower weights pretrained in [OpenCLIP](https://github.com/mlfoundations/open_clip) on LAION and fine-tuned on ImageNet-1k in `timm` by Ross Wightman.
14
 
15
  Please see related OpenCLIP model cards for more details on pretrain:
 
16
  * https://huggingface.co/laion/CLIP-convnext_large_d.laion2B-s26B-b102K-augreg
17
  * https://huggingface.co/laion/CLIP-convnext_base_w-laion2B-s13B-b82K-augreg
18
  * https://huggingface.co/laion/CLIP-convnext_base_w_320-laion_aesthetic-s13B-b82K
@@ -40,8 +41,9 @@ from urllib.request import urlopen
40
  from PIL import Image
41
  import timm
42
 
43
- img = Image.open(
44
- urlopen('https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'))
 
45
 
46
  model = timm.create_model('convnext_base.clip_laion2b_augreg_ft_in1k', pretrained=True)
47
  model = model.eval()
@@ -61,8 +63,9 @@ from urllib.request import urlopen
61
  from PIL import Image
62
  import timm
63
 
64
- img = Image.open(
65
- urlopen('https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'))
 
66
 
67
  model = timm.create_model(
68
  'convnext_base.clip_laion2b_augreg_ft_in1k',
@@ -79,11 +82,12 @@ output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batc
79
 
80
  for o in output:
81
  # print shape of each feature map in output
82
- # e.g. for convnext_base:
83
- # torch.Size([1, 128, 56, 56])
84
- # torch.Size([1, 256, 28, 28])
85
- # torch.Size([1, 512, 14, 14])
86
- # torch.Size([1, 1024, 7, 7])
 
87
  print(o.shape)
88
  ```
89
 
@@ -93,8 +97,9 @@ from urllib.request import urlopen
93
  from PIL import Image
94
  import timm
95
 
96
- img = Image.open(
97
- urlopen('https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'))
 
98
 
99
  model = timm.create_model(
100
  'convnext_base.clip_laion2b_augreg_ft_in1k',
@@ -112,122 +117,73 @@ output = model(transforms(img).unsqueeze(0)) # output is (batch_size, num_featu
112
  # or equivalently (without needing to set num_classes=0)
113
 
114
  output = model.forward_features(transforms(img).unsqueeze(0))
115
- # output is unpooled (ie.e a (batch_size, num_features, H, W) tensor
116
 
117
  output = model.forward_head(output, pre_logits=True)
118
- # output is (batch_size, num_features) tensor
119
  ```
120
 
121
  ## Model Comparison
122
- ### By Top-1
123
- All timing numbers from eager model PyTorch 1.13 on RTX 3090 w/ AMP.
124
-
125
- |model |top1 |top5 |img_size|param_count|gmacs |macts |samples_per_sec|batch_size|
126
- |----------------------------------------------|------|------|--------|-----------|------|------|---------------|----------|
127
- |[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 |
128
- |[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 |
129
- |[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 |
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_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 |
136
- |[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 |
137
- |[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 |
138
- |[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 |
139
- |[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 |
140
- |[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 |
141
- |[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 |
142
- |[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 |
143
- |[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 |
144
- |[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 |
145
- |[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 |
146
- |[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 |
147
- |[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 |
148
- |[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 |
149
- |[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 |
150
- |[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 |
151
- |[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 |
152
- |[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 |
153
- |[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 |
154
- |[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 |
155
- |[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 |
156
- |[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 |
157
- |[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 |
158
- |[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 |
159
- |[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 |
160
- |[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 |
161
- |[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 |
162
- |[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 |
163
- |[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 |
164
- |[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 |
165
- |[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 |
166
- |[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 |
167
- |[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 |
168
- |[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 |
169
- |[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 |
170
- |[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 |
171
- |[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 |
172
- |[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 |
173
- |[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 |
174
- |[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 |
175
- |[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 |
176
 
177
- ### By Throughput (samples / sec)
178
  All timing numbers from eager model PyTorch 1.13 on RTX 3090 w/ AMP.
179
 
180
- |model |top1 |top5 |img_size|param_count|gmacs |macts |samples_per_sec|batch_size|
181
- |----------------------------------------------|------|------|--------|-----------|------|------|---------------|----------|
182
- |[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 |
183
- |[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 |
184
- |[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 |
185
- |[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 |
186
- |[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 |
187
- |[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 |
188
- |[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 |
189
- |[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 |
190
- |[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 |
191
- |[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 |
192
- |[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 |
193
- |[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 |
194
- |[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 |
195
- |[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 |
196
- |[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 |
197
- |[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 |
198
- |[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 |
199
- |[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 |
200
- |[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 |
201
- |[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 |
202
- |[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 |
203
- |[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 |
204
- |[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 |
205
- |[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 |
206
- |[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 |
207
- |[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 |
208
- |[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 |
209
- |[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 |
210
- |[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 |
211
- |[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 |
212
- |[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 |
213
- |[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 |
214
- |[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 |
215
- |[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 |
216
- |[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 |
217
- |[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 |
218
- |[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 |
219
- |[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 |
220
- |[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 |
221
- |[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 |
222
- |[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 |
223
- |[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 |
224
- |[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 |
225
- |[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 |
226
- |[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 |
227
- |[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 |
228
- |[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 |
229
- |[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 |
230
- |[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 |
 
 
 
 
 
231
 
232
  ## Citation
233
  ```bibtex
@@ -287,7 +243,7 @@ All timing numbers from eager model PyTorch 1.13 on RTX 3090 w/ AMP.
287
  publisher = {GitHub},
288
  journal = {GitHub repository},
289
  doi = {10.5281/zenodo.4414861},
290
- howpublished = {\url{https://github.com/rwightman/pytorch-image-models}}
291
  }
292
  ```
293
  ```bibtex
 
13
  A ConvNeXt image classification model. CLIP image tower weights pretrained in [OpenCLIP](https://github.com/mlfoundations/open_clip) on LAION and fine-tuned on ImageNet-1k in `timm` by Ross Wightman.
14
 
15
  Please see related OpenCLIP model cards for more details on pretrain:
16
+ * https://huggingface.co/laion/CLIP-convnext_xxlarge-laion2B-s34B-b82K-augreg-soup
17
  * https://huggingface.co/laion/CLIP-convnext_large_d.laion2B-s26B-b102K-augreg
18
  * https://huggingface.co/laion/CLIP-convnext_base_w-laion2B-s13B-b82K-augreg
19
  * https://huggingface.co/laion/CLIP-convnext_base_w_320-laion_aesthetic-s13B-b82K
 
41
  from PIL import Image
42
  import timm
43
 
44
+ img = Image.open(urlopen(
45
+ 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'
46
+ ))
47
 
48
  model = timm.create_model('convnext_base.clip_laion2b_augreg_ft_in1k', pretrained=True)
49
  model = model.eval()
 
63
  from PIL import Image
64
  import timm
65
 
66
+ img = Image.open(urlopen(
67
+ 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'
68
+ ))
69
 
70
  model = timm.create_model(
71
  'convnext_base.clip_laion2b_augreg_ft_in1k',
 
82
 
83
  for o in output:
84
  # print shape of each feature map in output
85
+ # e.g.:
86
+ # torch.Size([1, 128, 64, 64])
87
+ # torch.Size([1, 256, 32, 32])
88
+ # torch.Size([1, 512, 16, 16])
89
+ # torch.Size([1, 1024, 8, 8])
90
+
91
  print(o.shape)
92
  ```
93
 
 
97
  from PIL import Image
98
  import timm
99
 
100
+ img = Image.open(urlopen(
101
+ 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'
102
+ ))
103
 
104
  model = timm.create_model(
105
  'convnext_base.clip_laion2b_augreg_ft_in1k',
 
117
  # or equivalently (without needing to set num_classes=0)
118
 
119
  output = model.forward_features(transforms(img).unsqueeze(0))
120
+ # output is unpooled, a (1, 1024, 8, 8) shaped tensor
121
 
122
  output = model.forward_head(output, pre_logits=True)
123
+ # output is a (1, num_features) shaped tensor
124
  ```
125
 
126
  ## Model Comparison
127
+ Explore the dataset and runtime metrics of this model in timm [model results](https://github.com/huggingface/pytorch-image-models/tree/main/results).
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
128
 
 
129
  All timing numbers from eager model PyTorch 1.13 on RTX 3090 w/ AMP.
130
 
131
+ | model |top1 |top5 |img_size|param_count|gmacs |macts |samples_per_sec|batch_size|
132
+ |------------------------------------------------------------------------------------------------------------------------------|------|------|--------|-----------|------|------|---------------|----------|
133
+ | [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 |
134
+ | [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 |
135
+ | [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 |
136
+ | [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 |
137
+ | [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 |
138
+ | [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 |
139
+ | [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 |
140
+ | [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 |
141
+ | [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 |
142
+ | [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 |
143
+ | [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 |
144
+ | [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 |
145
+ | [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 |
146
+ | [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 |
147
+ | [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 |
148
+ | [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 |
149
+ | [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 |
150
+ | [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 |
151
+ | [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 |
152
+ | [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 |
153
+ | [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 |
154
+ | [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 |
155
+ | [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 |
156
+ | [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 |
157
+ | [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 |
158
+ | [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 |
159
+ | [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 |
160
+ | [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 |
161
+ | [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 |
162
+ | [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 |
163
+ | [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 |
164
+ | [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 |
165
+ | [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 |
166
+ | [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 |
167
+ | [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 |
168
+ | [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 |
169
+ | [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 |
170
+ | [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 |
171
+ | [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 |
172
+ | [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 |
173
+ | [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 |
174
+ | [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 |
175
+ | [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 |
176
+ | [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 |
177
+ | [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 |
178
+ | [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 |
179
+ | [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 |
180
+ | [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 |
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_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 |
183
+ | [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 |
184
+ | [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 |
185
+ | [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 |
186
+ | [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 |
187
 
188
  ## Citation
189
  ```bibtex
 
243
  publisher = {GitHub},
244
  journal = {GitHub repository},
245
  doi = {10.5281/zenodo.4414861},
246
+ howpublished = {\url{https://github.com/huggingface/pytorch-image-models}}
247
  }
248
  ```
249
  ```bibtex
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c57160d267176d23eecd13575f28900c7781ac114638b781e7af7115faf60aa4
3
+ size 354400320