比較文本

尋找兩個文字檔案之間的差異

Real-time diff
Unified diff
Collapse lines
Highlight change
Syntax highlighting
工具
Diffchecker Desktop icon
Diffchecker Desktop The most secure way to run Diffchecker. Get the Diffchecker Desktop app: your diffs never leave your computer!Get Desktop

clip pruning without ignored layers

Created Diff never expires
72 removals
Lines
Total
Removed
Words
Total
Removed
To continue using this feature, upgrade to
Diffchecker logo
Diffchecker Pro
449 lines
73 additions
Lines
Total
Added
Words
Total
Added
To continue using this feature, upgrade to
Diffchecker logo
Diffchecker Pro
450 lines
model CLIP(
model CLIP(
(visual): VisionTransformer(
(visual): VisionTransformer(
(conv1): Conv2d(3, 1024, kernel_size=(14, 14), stride=(14, 14), bias=False)
(conv1): Conv2d(3, 1024, kernel_size=(14, 14), stride=(14, 14), bias=False)
(ln_pre): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_pre): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(transformer): Transformer(
(transformer): Transformer(
(resblocks): Sequential(
(resblocks): Sequential(
(0): ResidualAttentionBlock(
(0): ResidualAttentionBlock(
(attn): MultiheadAttention(
(attn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
)
)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(mlp): Sequential(
(mlp): Sequential(
(c_fc): Linear(in_features=1024, out_features=4096, bias=True)
(c_fc): Linear(in_features=1024, out_features=2048, bias=True)
(gelu): QuickGELU()
(gelu): QuickGELU()
(c_proj): Linear(in_features=4096, out_features=1024, bias=True)
(c_proj): Linear(in_features=2048, out_features=1024, bias=True)
)
)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
)
)
(1): ResidualAttentionBlock(
(1): ResidualAttentionBlock(
(attn): MultiheadAttention(
(attn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
)
)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(mlp): Sequential(
(mlp): Sequential(
(c_fc): Linear(in_features=1024, out_features=4096, bias=True)
(c_fc): Linear(in_features=1024, out_features=2048, bias=True)
(gelu): QuickGELU()
(gelu): QuickGELU()
(c_proj): Linear(in_features=4096, out_features=1024, bias=True)
(c_proj): Linear(in_features=2048, out_features=1024, bias=True)
)
)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
)
)
(2): ResidualAttentionBlock(
(2): ResidualAttentionBlock(
(attn): MultiheadAttention(
(attn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
)
)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(mlp): Sequential(
(mlp): Sequential(
(c_fc): Linear(in_features=1024, out_features=4096, bias=True)
(c_fc): Linear(in_features=1024, out_features=2048, bias=True)
(gelu): QuickGELU()
(gelu): QuickGELU()
(c_proj): Linear(in_features=4096, out_features=1024, bias=True)
(c_proj): Linear(in_features=2048, out_features=1024, bias=True)
)
)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
)
)
(3): ResidualAttentionBlock(
(3): ResidualAttentionBlock(
(attn): MultiheadAttention(
(attn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
)
)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(mlp): Sequential(
(mlp): Sequential(
(c_fc): Linear(in_features=1024, out_features=4096, bias=True)
(c_fc): Linear(in_features=1024, out_features=2048, bias=True)
(gelu): QuickGELU()
(gelu): QuickGELU()
(c_proj): Linear(in_features=4096, out_features=1024, bias=True)
(c_proj): Linear(in_features=2048, out_features=1024, bias=True)
)
)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
)
)
(4): ResidualAttentionBlock(
(4): ResidualAttentionBlock(
(attn): MultiheadAttention(
(attn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
)
)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(mlp): Sequential(
(mlp): Sequential(
(c_fc): Linear(in_features=1024, out_features=4096, bias=True)
(c_fc): Linear(in_features=1024, out_features=2048, bias=True)
(gelu): QuickGELU()
(gelu): QuickGELU()
(c_proj): Linear(in_features=4096, out_features=1024, bias=True)
(c_proj): Linear(in_features=2048, out_features=1024, bias=True)
)
)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
)
)
(5): ResidualAttentionBlock(
(5): ResidualAttentionBlock(
(attn): MultiheadAttention(
(attn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
)
)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(mlp): Sequential(
(mlp): Sequential(
(c_fc): Linear(in_features=1024, out_features=4096, bias=True)
(c_fc): Linear(in_features=1024, out_features=2048, bias=True)
(gelu): QuickGELU()
(gelu): QuickGELU()
(c_proj): Linear(in_features=4096, out_features=1024, bias=True)
(c_proj): Linear(in_features=2048, out_features=1024, bias=True)
)
)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
)
)
(6): ResidualAttentionBlock(
(6): ResidualAttentionBlock(
(attn): MultiheadAttention(
(attn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
)
)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(mlp): Sequential(
(mlp): Sequential(
(c_fc): Linear(in_features=1024, out_features=4096, bias=True)
(c_fc): Linear(in_features=1024, out_features=2048, bias=True)
(gelu): QuickGELU()
(gelu): QuickGELU()
(c_proj): Linear(in_features=4096, out_features=1024, bias=True)
(c_proj): Linear(in_features=2048, out_features=1024, bias=True)
)
)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
)
)
(7): ResidualAttentionBlock(
(7): ResidualAttentionBlock(
(attn): MultiheadAttention(
(attn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
)
)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(mlp): Sequential(
(mlp): Sequential(
(c_fc): Linear(in_features=1024, out_features=4096, bias=True)
(c_fc): Linear(in_features=1024, out_features=2048, bias=True)
(gelu): QuickGELU()
(gelu): QuickGELU()
(c_proj): Linear(in_features=4096, out_features=1024, bias=True)
(c_proj): Linear(in_features=2048, out_features=1024, bias=True)
)
)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
)
)
(8): ResidualAttentionBlock(
(8): ResidualAttentionBlock(
(attn): MultiheadAttention(
(attn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
)
)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(mlp): Sequential(
(mlp): Sequential(
(c_fc): Linear(in_features=1024, out_features=4096, bias=True)
(c_fc): Linear(in_features=1024, out_features=2048, bias=True)
(gelu): QuickGELU()
(gelu): QuickGELU()
(c_proj): Linear(in_features=4096, out_features=1024, bias=True)
(c_proj): Linear(in_features=2048, out_features=1024, bias=True)
)
)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
)
)
(9): ResidualAttentionBlock(
(9): ResidualAttentionBlock(
(attn): MultiheadAttention(
(attn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
)
)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(mlp): Sequential(
(mlp): Sequential(
(c_fc): Linear(in_features=1024, out_features=4096, bias=True)
(c_fc): Linear(in_features=1024, out_features=2048, bias=True)
(gelu): QuickGELU()
(gelu): QuickGELU()
(c_proj): Linear(in_features=4096, out_features=1024, bias=True)
(c_proj): Linear(in_features=2048, out_features=1024, bias=True)
)
)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
)
)
(10): ResidualAttentionBlock(
(10): ResidualAttentionBlock(
(attn): MultiheadAttention(
(attn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
)
)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(mlp): Sequential(
(mlp): Sequential(
(c_fc): Linear(in_features=1024, out_features=4096, bias=True)
(c_fc): Linear(in_features=1024, out_features=2048, bias=True)
(gelu): QuickGELU()
(gelu): QuickGELU()
(c_proj): Linear(in_features=4096, out_features=1024, bias=True)
(c_proj): Linear(in_features=2048, out_features=1024, bias=True)
)
)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
)
)
(11): ResidualAttentionBlock(
(11): ResidualAttentionBlock(
(attn): MultiheadAttention(
(attn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
)
)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(mlp): Sequential(
(mlp): Sequential(
(c_fc): Linear(in_features=1024, out_features=4096, bias=True)
(c_fc): Linear(in_features=1024, out_features=2048, bias=True)
(gelu): QuickGELU()
(gelu): QuickGELU()
(c_proj): Linear(in_features=4096, out_features=1024, bias=True)
(c_proj): Linear(in_features=2048, out_features=1024, bias=True)
)
)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
)
)
(12): ResidualAttentionBlock(
(12): ResidualAttentionBlock(
(attn): MultiheadAttention(
(attn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
)
)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(mlp): Sequential(
(mlp): Sequential(
(c_fc): Linear(in_features=1024, out_features=4096, bias=True)
(c_fc): Linear(in_features=1024, out_features=2048, bias=True)
(gelu): QuickGELU()
(gelu): QuickGELU()
(c_proj): Linear(in_features=4096, out_features=1024, bias=True)
(c_proj): Linear(in_features=2048, out_features=1024, bias=True)
)
)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
)
)
(13): ResidualAttentionBlock(
(13): ResidualAttentionBlock(
(attn): MultiheadAttention(
(attn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
)
)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(mlp): Sequential(
(mlp): Sequential(
(c_fc): Linear(in_features=1024, out_features=4096, bias=True)
(c_fc): Linear(in_features=1024, out_features=2048, bias=True)
(gelu): QuickGELU()
(gelu): QuickGELU()
(c_proj): Linear(in_features=4096, out_features=1024, bias=True)
(c_proj): Linear(in_features=2048, out_features=1024, bias=True)
)
)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
)
)
(14): ResidualAttentionBlock(
(14): ResidualAttentionBlock(
(attn): MultiheadAttention(
(attn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
)
)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(mlp): Sequential(
(mlp): Sequential(
(c_fc): Linear(in_features=1024, out_features=4096, bias=True)
(c_fc): Linear(in_features=1024, out_features=2048, bias=True)
(gelu): QuickGELU()
(gelu): QuickGELU()
(c_proj): Linear(in_features=4096, out_features=1024, bias=True)
(c_proj): Linear(in_features=2048, out_features=1024, bias=True)
)
)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
)
)
(15): ResidualAttentionBlock(
(15): ResidualAttentionBlock(
(attn): MultiheadAttention(
(attn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
)
)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(mlp): Sequential(
(mlp): Sequential(
(c_fc): Linear(in_features=1024, out_features=4096, bias=True)
(c_fc): Linear(in_features=1024, out_features=2048, bias=True)
(gelu): QuickGELU()
(gelu): QuickGELU()
(c_proj): Linear(in_features=4096, out_features=1024, bias=True)
(c_proj): Linear(in_features=2048, out_features=1024, bias=True)
)
)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
)
)
(16): ResidualAttentionBlock(
(16): ResidualAttentionBlock(
(attn): MultiheadAttention(
(attn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
)
)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(mlp): Sequential(
(mlp): Sequential(
(c_fc): Linear(in_features=1024, out_features=4096, bias=True)
(c_fc): Linear(in_features=1024, out_features=2048, bias=True)
(gelu): QuickGELU()
(gelu): QuickGELU()
(c_proj): Linear(in_features=4096, out_features=1024, bias=True)
(c_proj): Linear(in_features=2048, out_features=1024, bias=True)
)
)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
)
)
(17): ResidualAttentionBlock(
(17): ResidualAttentionBlock(
(attn): MultiheadAttention(
(attn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
)
)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(mlp): Sequential(
(mlp): Sequential(
(c_fc): Linear(in_features=1024, out_features=4096, bias=True)
(c_fc): Linear(in_features=1024, out_features=2048, bias=True)
(gelu): QuickGELU()
(gelu): QuickGELU()
(c_proj): Linear(in_features=4096, out_features=1024, bias=True)
(c_proj): Linear(in_features=2048, out_features=1024, bias=True)
)
)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
)
)
(18): ResidualAttentionBlock(
(18): ResidualAttentionBlock(
(attn): MultiheadAttention(
(attn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
)
)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(mlp): Sequential(
(mlp): Sequential(
(c_fc): Linear(in_features=1024, out_features=4096, bias=True)
(c_fc): Linear(in_features=1024, out_features=2048, bias=True)
(gelu): QuickGELU()
(gelu): QuickGELU()
(c_proj): Linear(in_features=4096, out_features=1024, bias=True)
(c_proj): Linear(in_features=2048, out_features=1024, bias=True)
)
)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
)
)
(19): ResidualAttentionBlock(
(19): ResidualAttentionBlock(
(attn): MultiheadAttention(
(attn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
)
)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(mlp): Sequential(
(mlp): Sequential(
(c_fc): Linear(in_features=1024, out_features=4096, bias=True)
(c_fc): Linear(in_features=1024, out_features=2048, bias=True)
(gelu): QuickGELU()
(gelu): QuickGELU()
(c_proj): Linear(in_features=4096, out_features=1024, bias=True)
(c_proj): Linear(in_features=2048, out_features=1024, bias=True)
)
)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
)
)
(20): ResidualAttentionBlock(
(20): ResidualAttentionBlock(
(attn): MultiheadAttention(
(attn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
)
)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(mlp): Sequential(
(mlp): Sequential(
(c_fc): Linear(in_features=1024, out_features=4096, bias=True)
(c_fc): Linear(in_features=1024, out_features=2048, bias=True)
(gelu): QuickGELU()
(gelu): QuickGELU()
(c_proj): Linear(in_features=4096, out_features=1024, bias=True)
(c_proj): Linear(in_features=2048, out_features=1024, bias=True)
)
)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
)
)
(21): ResidualAttentionBlock(
(21): ResidualAttentionBlock(
(attn): MultiheadAttention(
(attn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
)
)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(mlp): Sequential(
(mlp): Sequential(
(c_fc): Linear(in_features=1024, out_features=4096, bias=True)
(c_fc): Linear(in_features=1024, out_features=2048, bias=True)
(gelu): QuickGELU()
(gelu): QuickGELU()
(c_proj): Linear(in_features=4096, out_features=1024, bias=True)
(c_proj): Linear(in_features=2048, out_features=1024, bias=True)
)
)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
)
)
(22): ResidualAttentionBlock(
(22): ResidualAttentionBlock(
(attn): MultiheadAttention(
(attn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
)
)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(mlp): Sequential(
(mlp): Sequential(
(c_fc): Linear(in_features=1024, out_features=4096, bias=True)
(c_fc): Linear(in_features=1024, out_features=2048, bias=True)
(gelu): QuickGELU()
(gelu): QuickGELU()
(c_proj): Linear(in_features=4096, out_features=1024, bias=True)
(c_proj): Linear(in_features=2048, out_features=1024, bias=True)
)
)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
)
)
(23): ResidualAttentionBlock(
(23): ResidualAttentionBlock(
(attn): MultiheadAttention(
(attn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
(out_proj): NonDynamicallyQuantizableLinear(in_features=1024, out_features=1024, bias=True)
)
)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_1): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(mlp): Sequential(
(mlp): Sequential(
(c_fc): Linear(in_features=1024, out_features=4096, bias=True)
(c_fc): Linear(in_features=1024, out_features=2048, bias=True)
(gelu): QuickGELU()
(gelu): QuickGELU()
(c_proj): Linear(in_features=4096, out_features=1024, bias=True)
(c_proj): Linear(in_features=2048, out_features=1024, bias=True)
)
)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_2): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
)
)
)
)
)
)
(ln_post): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(ln_post): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
)
)
(transformer): Transformer(
(transformer): Transformer(
(resblocks): Sequential(
(resblocks): Sequential(
(0): ResidualAttentionBlock(
(0): ResidualAttentionBlock(
(attn): MultiheadAttention(
(attn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
(out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
)
)
(ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(mlp): Sequential(
(mlp): Sequential(
(c_fc): Linear(in_features=768, out_features=3072, bias=True)
(c_fc): Linear(in_features=768, out_features=1536, bias=True)
(gelu): QuickGELU()
(gelu): QuickGELU()
(c_proj): Linear(in_features=3072, out_features=768, bias=True)
(c_proj): Linear(in_features=1536, out_features=768, bias=True)
)
)
(ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
)
)
(1): ResidualAttentionBlock(
(1): ResidualAttentionBlock(
(attn): MultiheadAttention(
(attn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
(out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
)
)
(ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(mlp): Sequential(
(mlp): Sequential(
(c_fc): Linear(in_features=768, out_features=3072, bias=True)
(c_fc): Linear(in_features=768, out_features=1536, bias=True)
(gelu): QuickGELU()
(gelu): QuickGELU()
(c_proj): Linear(in_features=3072, out_features=768, bias=True)
(c_proj): Linear(in_features=1536, out_features=768, bias=True)
)
)
(ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
)
)
(2): ResidualAttentionBlock(
(2): ResidualAttentionBlock(
(attn): MultiheadAttention(
(attn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
(out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
)
)
(ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(mlp): Sequential(
(mlp): Sequential(
(c_fc): Linear(in_features=768, out_features=3072, bias=True)
(c_fc): Linear(in_features=768, out_features=1536, bias=True)
(gelu): QuickGELU()
(gelu): QuickGELU()
(c_proj): Linear(in_features=3072, out_features=768, bias=True)
(c_proj): Linear(in_features=1536, out_features=768, bias=True)
)
)
(ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
)
)
(3): ResidualAttentionBlock(
(3): ResidualAttentionBlock(
(attn): MultiheadAttention(
(attn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
(out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
)
)
(ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(mlp): Sequential(
(mlp): Sequential(
(c_fc): Linear(in_features=768, out_features=3072, bias=True)
(c_fc): Linear(in_features=768, out_features=1536, bias=True)
(gelu): QuickGELU()
(gelu): QuickGELU()
(c_proj): Linear(in_features=3072, out_features=768, bias=True)
(c_proj): Linear(in_features=1536, out_features=768, bias=True)
)
)
(ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
)
)
(4): ResidualAttentionBlock(
(4): ResidualAttentionBlock(
(attn): MultiheadAttention(
(attn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
(out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
)
)
(ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(mlp): Sequential(
(mlp): Sequential(
(c_fc): Linear(in_features=768, out_features=3072, bias=True)
(c_fc): Linear(in_features=768, out_features=1536, bias=True)
(gelu): QuickGELU()
(gelu): QuickGELU()
(c_proj): Linear(in_features=3072, out_features=768, bias=True)
(c_proj): Linear(in_features=1536, out_features=768, bias=True)
)
)
(ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
)
)
(5): ResidualAttentionBlock(
(5): ResidualAttentionBlock(
(attn): MultiheadAttention(
(attn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
(out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
)
)
(ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(mlp): Sequential(
(mlp): Sequential(
(c_fc): Linear(in_features=768, out_features=3072, bias=True)
(c_fc): Linear(in_features=768, out_features=1536, bias=True)
(gelu): QuickGELU()
(gelu): QuickGELU()
(c_proj): Linear(in_features=3072, out_features=768, bias=True)
(c_proj): Linear(in_features=1536, out_features=768, bias=True)
)
)
(ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
)
)
(6): ResidualAttentionBlock(
(6): ResidualAttentionBlock(
(attn): MultiheadAttention(
(attn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
(out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
)
)
(ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(mlp): Sequential(
(mlp): Sequential(
(c_fc): Linear(in_features=768, out_features=3072, bias=True)
(c_fc): Linear(in_features=768, out_features=1536, bias=True)
(gelu): QuickGELU()
(gelu): QuickGELU()
(c_proj): Linear(in_features=3072, out_features=768, bias=True)
(c_proj): Linear(in_features=1536, out_features=768, bias=True)
)
)
(ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
)
)
(7): ResidualAttentionBlock(
(7): ResidualAttentionBlock(
(attn): MultiheadAttention(
(attn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
(out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
)
)
(ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(mlp): Sequential(
(mlp): Sequential(
(c_fc): Linear(in_features=768, out_features=3072, bias=True)
(c_fc): Linear(in_features=768, out_features=1536, bias=True)
(gelu): QuickGELU()
(gelu): QuickGELU()
(c_proj): Linear(in_features=3072, out_features=768, bias=True)
(c_proj): Linear(in_features=1536, out_features=768, bias=True)
)
)
(ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
)
)
(8): ResidualAttentionBlock(
(8): ResidualAttentionBlock(
(attn): MultiheadAttention(
(attn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
(out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
)
)
(ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(mlp): Sequential(
(mlp): Sequential(
(c_fc): Linear(in_features=768, out_features=3072, bias=True)
(c_fc): Linear(in_features=768, out_features=1536, bias=True)
(gelu): QuickGELU()
(gelu): QuickGELU()
(c_proj): Linear(in_features=3072, out_features=768, bias=True)
(c_proj): Linear(in_features=1536, out_features=768, bias=True)
)
)
(ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
)
)
(9): ResidualAttentionBlock(
(9): ResidualAttentionBlock(
(attn): MultiheadAttention(
(attn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
(out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
)
)
(ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(mlp): Sequential(
(mlp): Sequential(
(c_fc): Linear(in_features=768, out_features=3072, bias=True)
(c_fc): Linear(in_features=768, out_features=1536, bias=True)
(gelu): QuickGELU()
(gelu): QuickGELU()
(c_proj): Linear(in_features=3072, out_features=768, bias=True)
(c_proj): Linear(in_features=1536, out_features=768, bias=True)
)
)
(ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
)
)
(10): ResidualAttentionBlock(
(10): ResidualAttentionBlock(
(attn): MultiheadAttention(
(attn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
(out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
)
)
(ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(mlp): Sequential(
(mlp): Sequential(
(c_fc): Linear(in_features=768, out_features=3072, bias=True)
(c_fc): Linear(in_features=768, out_features=1536, bias=True)
(gelu): QuickGELU()
(gelu): QuickGELU()
(c_proj): Linear(in_features=3072, out_features=768, bias=True)
(c_proj): Linear(in_features=1536, out_features=768, bias=True)
)
)
(ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
)
)
(11): ResidualAttentionBlock(
(11): ResidualAttentionBlock(
(attn): MultiheadAttention(
(attn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
(out_proj): NonDynamicallyQuantizableLinear(in_features=768, out_features=768, bias=True)
)
)
(ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(ln_1): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(mlp): Sequential(
(mlp): Sequential(
(c_fc): Linear(in_features=768, out_features=3072, bias=True)
(c_fc): Linear(in_features=768, out_features=1536, bias=True)
(gelu): QuickGELU()
(gelu): QuickGELU()
(c_proj): Linear(in_features=3072, out_features=768, bias=True)
(c_proj): Linear(in_features=1536, out_features=768, bias=True)
)
)
(ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(ln_2): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
)
)
)
)
)
)
(token_embedding): Embedding(49408, 768)
(token_embedding): Embedding(49408, 768)
(ln_final): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
(ln_final): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
)
)

已保存差異