Initial commit

open_clip/src/open_clip/model_configs/ViT-L-14-336.json

+{
+    "embed_dim": 768,
+    "vision_cfg": {
+        "image_size": 336,
+        "layers": 24,
+        "width": 1024,
+        "patch_size": 14
+    },
+    "text_cfg": {
+        "context_length": 77,
+        "vocab_size": 49408,
+        "width": 768,
+        "heads": 12,
+        "layers": 12
+    }
+}
\ No newline at end of file

open_clip/src/open_clip/model_configs/ViT-L-14.json

+{
+    "embed_dim": 768,
+    "vision_cfg": {
+        "image_size": 224,
+        "layers": 24,
+        "width": 1024,
+        "patch_size": 14
+    },
+    "text_cfg": {
+        "context_length": 77,
+        "vocab_size": 49408,
+        "width": 768,
+        "heads": 12,
+        "layers": 12
+    }
+}
\ No newline at end of file

open_clip/src/open_clip/model_configs/ViT-L-16-320.json

+{
+    "embed_dim": 768,
+    "vision_cfg": {
+        "image_size": 320,
+        "layers": 24,
+        "width": 1024,
+        "patch_size": 16
+    },
+    "text_cfg": {
+        "context_length": 77,
+        "vocab_size": 49408,
+        "width": 768,
+        "heads": 12,
+        "layers": 12
+    }
+}
\ No newline at end of file

open_clip/src/open_clip/model_configs/ViT-L-16.json

+{
+    "embed_dim": 768,
+    "vision_cfg": {
+        "image_size": 224,
+        "layers": 24,
+        "width": 1024,
+        "patch_size": 16
+    },
+    "text_cfg": {
+        "context_length": 77,
+        "vocab_size": 49408,
+        "width": 768,
+        "heads": 12,
+        "layers": 12
+    }
+}
\ No newline at end of file

open_clip/src/open_clip/model_configs/ViT-M-16-alt.json

+{
+    "embed_dim": 384,
+    "vision_cfg": {
+        "image_size": 224,
+        "layers": 12,
+        "width": 512,
+        "patch_size": 16,
+        "ls_init_value": 1e-4
+    },
+    "text_cfg": {
+        "context_length": 77,
+        "vocab_size": 49408,
+        "width": 384,
+        "heads": 6,
+        "layers": 12
+    }
+}
\ No newline at end of file

open_clip/src/open_clip/model_configs/ViT-M-16.json

+{
+    "embed_dim": 512,
+    "vision_cfg": {
+        "image_size": 224,
+        "layers": 12,
+        "width": 512,
+        "patch_size": 16
+    },
+    "text_cfg": {
+        "context_length": 77,
+        "vocab_size": 49408,
+        "width": 512,
+        "heads": 8,
+        "layers": 12
+    }
+}
\ No newline at end of file

open_clip/src/open_clip/model_configs/ViT-M-32-alt.json

+{
+    "embed_dim": 384,
+    "vision_cfg": {
+        "image_size": 224,
+        "layers": 12,
+        "width": 512,
+        "patch_size": 32
+    },
+    "text_cfg": {
+        "context_length": 77,
+        "vocab_size": 49408,
+        "width": 384,
+        "heads": 6,
+        "layers": 12
+    }
+}
\ No newline at end of file

open_clip/src/open_clip/model_configs/ViT-M-32.json

+{
+    "embed_dim": 512,
+    "vision_cfg": {
+        "image_size": 224,
+        "layers": 12,
+        "width": 512,
+        "patch_size": 32
+    },
+    "text_cfg": {
+        "context_length": 77,
+        "vocab_size": 49408,
+        "width": 512,
+        "heads": 8,
+        "layers": 12
+    }
+}
\ No newline at end of file

open_clip/src/open_clip/model_configs/ViT-S-16-alt.json

+{
+    "embed_dim": 256,
+    "vision_cfg": {
+        "image_size": 224,
+        "layers": 12,
+        "width": 384,
+        "patch_size": 16
+    },
+    "text_cfg": {
+        "context_length": 77,
+        "vocab_size": 49408,
+        "width": 256,
+        "heads": 4,
+        "layers": 10
+    }
+}
\ No newline at end of file

open_clip/src/open_clip/model_configs/ViT-S-16.json

+{
+    "embed_dim": 384,
+    "vision_cfg": {
+        "image_size": 224,
+        "layers": 12,
+        "width": 384,
+        "patch_size": 16
+    },
+    "text_cfg": {
+        "context_length": 77,
+        "vocab_size": 49408,
+        "width": 384,
+        "heads": 6,
+        "layers": 12
+    }
+}
\ No newline at end of file

open_clip/src/open_clip/model_configs/ViT-S-32-alt.json

+{
+    "embed_dim": 256,
+    "vision_cfg": {
+        "image_size": 224,
+        "layers": 12,
+        "width": 384,
+        "patch_size": 32
+    },
+    "text_cfg": {
+        "context_length": 77,
+        "vocab_size": 49408,
+        "width": 256,
+        "heads": 4,
+        "layers": 10
+    }
+}
\ No newline at end of file

open_clip/src/open_clip/model_configs/ViT-S-32.json

+{
+    "embed_dim": 384,
+    "vision_cfg": {
+        "image_size": 224,
+        "layers": 12,
+        "width": 384,
+        "patch_size": 32
+    },
+    "text_cfg": {
+        "context_length": 77,
+        "vocab_size": 49408,
+        "width": 384,
+        "heads": 6,
+        "layers": 12
+    }
+}
\ No newline at end of file

open_clip/src/open_clip/model_configs/ViT-bigG-14.json

+{
+    "embed_dim": 1280,
+    "vision_cfg": {
+        "image_size": 224,
+        "layers": 48,
+        "width": 1664,
+        "head_width": 104,
+        "mlp_ratio": 4.9231,
+        "patch_size": 14
+    },
+    "text_cfg": {
+        "context_length": 77,
+        "vocab_size": 49408,
+        "width": 1280,
+        "heads": 20,
+        "layers": 32
+    }
+}
\ No newline at end of file

open_clip/src/open_clip/model_configs/ViT-e-14.json

+{
+    "embed_dim": 1280,
+    "vision_cfg": {
+        "image_size": 224,
+        "layers": 56,
+        "width": 1792,
+        "head_width": 112,
+        "mlp_ratio": 8.5715,
+        "patch_size": 14
+    },
+    "text_cfg": {
+        "context_length": 77,
+        "vocab_size": 49408,
+        "width": 1280,
+        "heads": 20,
+        "layers": 36
+    }
+}
\ No newline at end of file

open_clip/src/open_clip/model_configs/ViT-g-14.json

+{
+    "embed_dim": 1024,
+    "vision_cfg": {
+        "image_size": 224,
+        "layers": 40,
+        "width": 1408,
+        "head_width": 88,
+        "mlp_ratio": 4.3637,
+        "patch_size": 14
+    },
+    "text_cfg": {
+        "context_length": 77,
+        "vocab_size": 49408,
+        "width": 1024,
+        "heads": 16,
+        "layers": 24
+    }
+}
\ No newline at end of file

open_clip/src/open_clip/model_configs/convnext_base.json

+{
+    "embed_dim": 512,
+    "vision_cfg": {
+        "timm_model_name": "convnext_base",
+        "timm_model_pretrained": false,
+        "timm_pool": "",
+        "timm_proj": "linear",
+        "image_size": 224
+    },
+    "text_cfg": {
+        "context_length": 77,
+        "vocab_size": 49408,
+        "width": 512,
+        "heads": 8,
+        "layers": 12
+    }
+}
\ No newline at end of file

open_clip/src/open_clip/model_configs/convnext_base_w.json

+{
+    "embed_dim": 640,
+    "vision_cfg": {
+        "timm_model_name": "convnext_base",
+        "timm_model_pretrained": false,
+        "timm_pool": "",
+        "timm_proj": "linear",
+        "image_size": 256
+    },
+    "text_cfg": {
+        "context_length": 77,
+        "vocab_size": 49408,
+        "width": 640,
+        "heads": 10,
+        "layers": 12
+    }
+}
\ No newline at end of file

open_clip/src/open_clip/model_configs/convnext_base_w_320.json

+{
+    "embed_dim": 640,
+    "vision_cfg": {
+        "timm_model_name": "convnext_base",
+        "timm_model_pretrained": false,
+        "timm_pool": "",
+        "timm_proj": "linear",
+        "image_size": 320
+    },
+    "text_cfg": {
+        "context_length": 77,
+        "vocab_size": 49408,
+        "width": 640,
+        "heads": 10,
+        "layers": 12
+    }
+}
\ No newline at end of file

open_clip/src/open_clip/model_configs/convnext_large.json

+{
+    "embed_dim": 768,
+    "vision_cfg": {
+        "timm_model_name": "convnext_large",
+        "timm_model_pretrained": false,
+        "timm_pool": "",
+        "timm_proj": "linear",
+        "image_size": 224
+    },
+    "text_cfg": {
+        "context_length": 77,
+        "vocab_size": 49408,
+        "width": 768,
+        "heads": 12,
+        "layers": 12
+    }
+}
\ No newline at end of file

open_clip/src/open_clip/model_configs/convnext_large_d.json

+{
+    "embed_dim": 768,
+    "vision_cfg": {
+        "timm_model_name": "convnext_large",
+        "timm_model_pretrained": false,
+        "timm_pool": "",
+        "timm_proj": "mlp",
+        "timm_drop": 0.1,
+        "timm_drop_path": 0.1,
+        "image_size": 256
+    },
+    "text_cfg": {
+        "context_length": 77,
+        "vocab_size": 49408,
+        "width": 768,
+        "heads": 12,
+        "layers": 16
+    }
+}
\ No newline at end of file