Sigh :)
35a5639
verified
-
1.52 kB
initial commit
-
4.79 kB
Add dataset info to benchmark results
-
931 MB
Add HuggingFace format full model + text encoder only
-
323 MB
Add HuggingFace format full model + text encoder only
ViT-L-14-BEST-smooth-GmP-ft-pickle-OpenAI.pt
Detected Pickle imports (18)
- "torch._utils._rebuild_parameter",
- "clip.model.VisionTransformer",
- "clip.model.LayerNorm",
- "torch._utils._rebuild_tensor_v2",
- "__builtin__.set",
- "torch.nn.modules.activation.MultiheadAttention",
- "torch.nn.modules.linear.Linear",
- "torch.nn.modules.sparse.Embedding",
- "torch.HalfStorage",
- "torch.nn.modules.container.Sequential",
- "collections.OrderedDict",
- "torch.FloatStorage",
- "clip.model.ResidualAttentionBlock",
- "torch.nn.modules.linear.NonDynamicallyQuantizableLinear",
- "clip.model.Transformer",
- "clip.model.CLIP",
- "torch.nn.modules.conv.Conv2d",
- "clip.model.QuickGELU"
How to fix it?
932 MB
Add new #1 top-performing smooth-GmP model
-
932 MB
Add new #1 top-performing smooth-GmP model
-
932 MB
Add new #1 top-performing smooth-GmP model
-
325 MB
ViT-L Text Encoder "same as HF / used in SDXL"
ViT-L-14-GmP-ft-pickle-OpenAI.pt
Detected Pickle imports (18)
- "torch._utils._rebuild_parameter",
- "clip.model.VisionTransformer",
- "clip.model.LayerNorm",
- "torch._utils._rebuild_tensor_v2",
- "__builtin__.set",
- "torch.nn.modules.activation.MultiheadAttention",
- "torch.nn.modules.linear.Linear",
- "torch.nn.modules.sparse.Embedding",
- "torch.HalfStorage",
- "torch.nn.modules.container.Sequential",
- "collections.OrderedDict",
- "torch.FloatStorage",
- "clip.model.ResidualAttentionBlock",
- "torch.nn.modules.linear.NonDynamicallyQuantizableLinear",
- "clip.model.Transformer",
- "clip.model.CLIP",
- "torch.nn.modules.conv.Conv2d",
- "clip.model.QuickGELU"
How to fix it?
932 MB
Added .safetensors and full model object pickle
-
1.71 GB
Upload ViT-L-14-GmP-ft-state_dict.pt
-
932 MB
Added .safetensors and full model object pickle
-
931 MB
Greatly improved TEXT + Detail (as CLIP-L for Flux.1)
-
323 MB
Greatly improved TEXT + Detail (as CLIP-L for Flux.1)
ViT-L-14-TEXT-detail-improved-hiT-GmP-pickle-OpenAI.pt
Detected Pickle imports (18)
- "clip.model.VisionTransformer",
- "torch.HalfStorage",
- "__builtin__.set",
- "torch.nn.modules.linear.NonDynamicallyQuantizableLinear",
- "torch.nn.modules.linear.Linear",
- "clip.model.Transformer",
- "torch.FloatStorage",
- "clip.model.ResidualAttentionBlock",
- "clip.model.CLIP",
- "collections.OrderedDict",
- "torch._utils._rebuild_tensor_v2",
- "torch.nn.modules.activation.MultiheadAttention",
- "clip.model.QuickGELU",
- "torch.nn.modules.container.Sequential",
- "clip.model.LayerNorm",
- "torch.nn.modules.conv.Conv2d",
- "torch.nn.modules.sparse.Embedding",
- "torch._utils._rebuild_parameter"
How to fix it?
932 MB
Greatly improved TEXT + Detail (as CLIP-L for Flux.1)
-
932 MB
Greatly improved TEXT + Detail (as CLIP-L for Flux.1)
-
4.51 kB
Trying to just use the OpenAI config file here
-
525 kB
Sigh :)
-
931 MB
REDUNDANT version of the "TEXT" model. This is NOT a new model.
-
316 Bytes
From openai/clip-vit-large-patch14
-
389 Bytes
From openai/clip-vit-large-patch14
-
2.22 MB
From openai/clip-vit-large-patch14
-
902 Bytes
From openai/clip-vit-large-patch14
-
961 kB
From openai/clip-vit-large-patch14
