<
   ir_version: 10,
   opset_import: ["" : 23],
   producer_name: "pytorch"
>
main_graph (int32[batch,64] text) => (float[batch,768] text_embedding) 
   <
      float[batch,64,768] add_1071
      float[batch,64,768] add_1092
      float[batch,64,768] add_119
      float[batch,64,768] add_1207
      float[batch,64,768] add_1228
      float[batch,64,768] add_1343
      float[batch,64,768] add_1364
      float[batch,64,768] add_140
      float[batch,64,768] add_1479
      float[batch,64,768] add_1500
      float[batch,1,768] add_1500_pooled
      float[batch,1,768] add_1615
      float[batch,1,768] add_1636
      float[batch,64,768] add_255
      float[batch,64,768] add_276
      float[batch,64,768] add_391
      float[batch,64,768] add_4
      float[batch,64,768] add_412
      float[batch,64,768] add_527
      float[batch,64,768] add_548
      float[batch,64,768] add_663
      float[batch,64,768] add_684
      float[batch,64,768] add_799
      float[batch,64,768] add_820
      float[batch,64,768] add_935
      float[batch,64,768] add_956
      float[batch,1] clamp_min
      float[batch,64,768] embedding
      float[batch,64,3072] gelu
      float[batch,64,3072] gelu_1
      float[batch,64,3072] gelu_10
      float[batch,1,3072] gelu_11
      float[batch,64,3072] gelu_2
      float[batch,64,3072] gelu_3
      float[batch,64,3072] gelu_4
      float[batch,64,3072] gelu_5
      float[batch,64,3072] gelu_6
      float[batch,64,3072] gelu_7
      float[batch,64,3072] gelu_8
      float[batch,64,3072] gelu_9
      float[batch,64,768] layer_norm
      float[batch,64,768] layer_norm_1
      float[batch,64,768] layer_norm_10
      float[batch,64,768] layer_norm_11
      float[batch,64,768] layer_norm_12
      float[batch,64,768] layer_norm_13
      float[batch,64,768] layer_norm_14
      float[batch,64,768] layer_norm_15
      float[batch,64,768] layer_norm_16
      float[batch,64,768] layer_norm_17
      float[batch,64,768] layer_norm_18
      float[batch,64,768] layer_norm_19
      float[batch,64,768] layer_norm_2
      float[batch,64,768] layer_norm_20
      float[batch,64,768] layer_norm_21
      float[batch,64,768] layer_norm_22
      float[batch,1,768] layer_norm_23
      float[batch,64,768] layer_norm_3
      float[batch,64,768] layer_norm_4
      float[batch,64,768] layer_norm_5
      float[batch,64,768] layer_norm_6
      float[batch,64,768] layer_norm_7
      float[batch,64,768] layer_norm_8
      float[batch,64,768] layer_norm_9
      float[batch,1] linalg_vector_norm
      float[batch,64,3072] linear_10
      float[batch,64,768] linear_11
      float[batch,64,3072] linear_14
      float[batch,64,768] linear_15
      float[batch,64,3072] linear_18
      float[batch,64,768] linear_19
      float[batch,64,3072] linear_2
      float[batch,64,3072] linear_22
      float[batch,64,768] linear_23
      float[batch,64,3072] linear_26
      float[batch,64,768] linear_27
      float[batch,64,768] linear_3
      float[batch,64,3072] linear_30
      float[batch,64,768] linear_31
      float[batch,64,3072] linear_34
      float[batch,64,768] linear_35
      float[batch,64,3072] linear_38
      float[batch,64,768] linear_39
      float[batch,64,3072] linear_42
      float[batch,64,768] linear_43
      float[batch,1,3072] linear_46
      float[batch,1,768] linear_47
      float[batch,768] linear_48
      float[batch,64,3072] linear_6
      float[batch,64,768] linear_7
      float[batch,64,768] node_scaled_dot_product_attention_10_k
      float[batch,64,768] node_scaled_dot_product_attention_10_out
      float[batch,64,768] node_scaled_dot_product_attention_10_out_mm_out
      float[batch,64,768] node_scaled_dot_product_attention_10_q
      float[batch,64,2304] node_scaled_dot_product_attention_10_qkv
      float[batch,64,2304] node_scaled_dot_product_attention_10_qkv_mm_out
      float[batch,64,768] node_scaled_dot_product_attention_10_v
      float[batch,64,768] node_scaled_dot_product_attention_11_k
      float[batch,1,768] node_scaled_dot_product_attention_11_out
      float[batch,1,768] node_scaled_dot_product_attention_11_out_mm_out
      float[batch,64,768] node_scaled_dot_product_attention_11_q
      float[batch,1,768] node_scaled_dot_product_attention_11_q_pooled
      float[batch,64,2304] node_scaled_dot_product_attention_11_qkv
      float[batch,64,2304] node_scaled_dot_product_attention_11_qkv_mm_out
      float[batch,64,768] node_scaled_dot_product_attention_11_v
      float[batch,64,768] node_scaled_dot_product_attention_1_k
      float[batch,64,768] node_scaled_dot_product_attention_1_out
      float[batch,64,768] node_scaled_dot_product_attention_1_out_mm_out
      float[batch,64,768] node_scaled_dot_product_attention_1_q
      float[batch,64,2304] node_scaled_dot_product_attention_1_qkv
      float[batch,64,2304] node_scaled_dot_product_attention_1_qkv_mm_out
      float[batch,64,768] node_scaled_dot_product_attention_1_v
      float[batch,64,768] node_scaled_dot_product_attention_2_k
      float[batch,64,768] node_scaled_dot_product_attention_2_out
      float[batch,64,768] node_scaled_dot_product_attention_2_out_mm_out
      float[batch,64,768] node_scaled_dot_product_attention_2_q
      float[batch,64,2304] node_scaled_dot_product_attention_2_qkv
      float[batch,64,2304] node_scaled_dot_product_attention_2_qkv_mm_out
      float[batch,64,768] node_scaled_dot_product_attention_2_v
      float[batch,64,768] node_scaled_dot_product_attention_3_k
      float[batch,64,768] node_scaled_dot_product_attention_3_out
      float[batch,64,768] node_scaled_dot_product_attention_3_out_mm_out
      float[batch,64,768] node_scaled_dot_product_attention_3_q
      float[batch,64,2304] node_scaled_dot_product_attention_3_qkv
      float[batch,64,2304] node_scaled_dot_product_attention_3_qkv_mm_out
      float[batch,64,768] node_scaled_dot_product_attention_3_v
      float[batch,64,768] node_scaled_dot_product_attention_4_k
      float[batch,64,768] node_scaled_dot_product_attention_4_out
      float[batch,64,768] node_scaled_dot_product_attention_4_out_mm_out
      float[batch,64,768] node_scaled_dot_product_attention_4_q
      float[batch,64,2304] node_scaled_dot_product_attention_4_qkv
      float[batch,64,2304] node_scaled_dot_product_attention_4_qkv_mm_out
      float[batch,64,768] node_scaled_dot_product_attention_4_v
      float[batch,64,768] node_scaled_dot_product_attention_5_k
      float[batch,64,768] node_scaled_dot_product_attention_5_out
      float[batch,64,768] node_scaled_dot_product_attention_5_out_mm_out
      float[batch,64,768] node_scaled_dot_product_attention_5_q
      float[batch,64,2304] node_scaled_dot_product_attention_5_qkv
      float[batch,64,2304] node_scaled_dot_product_attention_5_qkv_mm_out
      float[batch,64,768] node_scaled_dot_product_attention_5_v
      float[batch,64,768] node_scaled_dot_product_attention_6_k
      float[batch,64,768] node_scaled_dot_product_attention_6_out
      float[batch,64,768] node_scaled_dot_product_attention_6_out_mm_out
      float[batch,64,768] node_scaled_dot_product_attention_6_q
      float[batch,64,2304] node_scaled_dot_product_attention_6_qkv
      float[batch,64,2304] node_scaled_dot_product_attention_6_qkv_mm_out
      float[batch,64,768] node_scaled_dot_product_attention_6_v
      float[batch,64,768] node_scaled_dot_product_attention_7_k
      float[batch,64,768] node_scaled_dot_product_attention_7_out
      float[batch,64,768] node_scaled_dot_product_attention_7_out_mm_out
      float[batch,64,768] node_scaled_dot_product_attention_7_q
      float[batch,64,2304] node_scaled_dot_product_attention_7_qkv
      float[batch,64,2304] node_scaled_dot_product_attention_7_qkv_mm_out
      float[batch,64,768] node_scaled_dot_product_attention_7_v
      float[batch,64,768] node_scaled_dot_product_attention_8_k
      float[batch,64,768] node_scaled_dot_product_attention_8_out
      float[batch,64,768] node_scaled_dot_product_attention_8_out_mm_out
      float[batch,64,768] node_scaled_dot_product_attention_8_q
      float[batch,64,2304] node_scaled_dot_product_attention_8_qkv
      float[batch,64,2304] node_scaled_dot_product_attention_8_qkv_mm_out
      float[batch,64,768] node_scaled_dot_product_attention_8_v
      float[batch,64,768] node_scaled_dot_product_attention_9_k
      float[batch,64,768] node_scaled_dot_product_attention_9_out
      float[batch,64,768] node_scaled_dot_product_attention_9_out_mm_out
      float[batch,64,768] node_scaled_dot_product_attention_9_q
      float[batch,64,2304] node_scaled_dot_product_attention_9_qkv
      float[batch,64,2304] node_scaled_dot_product_attention_9_qkv_mm_out
      float[batch,64,768] node_scaled_dot_product_attention_9_v
      float[batch,64,768] node_scaled_dot_product_attention_k
      float[batch,64,768] node_scaled_dot_product_attention_out
      float[batch,64,768] node_scaled_dot_product_attention_out_mm_out
      float[batch,64,768] node_scaled_dot_product_attention_q
      float[batch,64,2304] node_scaled_dot_product_attention_qkv
      float[batch,64,2304] node_scaled_dot_product_attention_qkv_mm_out
      float[batch,64,768] node_scaled_dot_product_attention_v
      float[batch,64,768] scaled_dot_product_attention
      float[batch,64,768] scaled_dot_product_attention_1
      float[batch,64,768] scaled_dot_product_attention_10
      float[batch,1,768] scaled_dot_product_attention_11
      float[batch,64,768] scaled_dot_product_attention_2
      float[batch,64,768] scaled_dot_product_attention_3
      float[batch,64,768] scaled_dot_product_attention_4
      float[batch,64,768] scaled_dot_product_attention_5
      float[batch,64,768] scaled_dot_product_attention_6
      float[batch,64,768] scaled_dot_product_attention_7
      float[batch,64,768] scaled_dot_product_attention_8
      float[batch,64,768] scaled_dot_product_attention_9
      float[batch,768] select_36
      float[batch,64,3072] val_41
      float[batch,64,768] val_42
      float[batch,64,3072] val_43
      float[batch,64,768] val_44
      float[batch,64,3072] val_45
      float[batch,64,768] val_46
      float[batch,64,3072] val_47
      float[batch,64,768] val_48
      float[batch,64,3072] val_49
      float[batch,64,768] val_50
      float[batch,64,3072] val_51
      float[batch,64,768] val_52
      float[batch,64,3072] val_53
      float[batch,64,768] val_54
      float[batch,64,3072] val_55
      float[batch,64,768] val_56
      float[batch,64,3072] val_57
      float[batch,64,768] val_58
      float[batch,64,3072] val_59
      float[batch,64,768] val_60
      float[batch,64,3072] val_61
      float[batch,64,768] val_62
      float[batch,1,3072] val_63
      float[batch,1,768] val_64
      float[batch,768] val_65
   >
{
   val_40 = Gather <axis: int = 0> ("text.token_embedding.weight_fp16", text)
   embedding = Cast <to: int = 1> (val_40)
   add_4 = Add (embedding, "text.positional_embedding")
   [node_layer_norm] layer_norm = LayerNormalization <axis: int = -1, epsilon: float = 1e-06, stash_type: int = 1> (add_4, "text.transformer.resblocks.0.ln_1.weight", "text.transformer.resblocks.0.ln_1.bias")
   [node_scaled_dot_product_attention_qkv_mm] node_scaled_dot_product_attention_qkv_mm_out = MatMul (layer_norm, val_4)
   [node_scaled_dot_product_attention_qkv_bias] node_scaled_dot_product_attention_qkv = Add (node_scaled_dot_product_attention_qkv_mm_out, "text.transformer.resblocks.0.attn.in_proj_bias")
   [node_scaled_dot_product_attention_qkv_split] node_scaled_dot_product_attention_q, node_scaled_dot_product_attention_k, node_scaled_dot_product_attention_v = Split <axis: int = -1> (node_scaled_dot_product_attention_qkv, attn3d_split_3x768)
   [node_scaled_dot_product_attention] scaled_dot_product_attention = Attention <is_causal: int = 0, kv_num_heads: int = 12, q_num_heads: int = 12, qk_matmul_output_mode: int = 0, softcap: float = 0> (node_scaled_dot_product_attention_q, node_scaled_dot_product_attention_k, node_scaled_dot_product_attention_v)
   [node_scaled_dot_product_attention_out_mm] node_scaled_dot_product_attention_out_mm_out = MatMul (scaled_dot_product_attention, node_scaled_dot_product_attention_wo_t)
   [node_scaled_dot_product_attention_out_bias] node_scaled_dot_product_attention_out = Add (node_scaled_dot_product_attention_out_mm_out, "text.transformer.resblocks.0.attn.out_proj.bias")
   add_119 = Add (add_4, node_scaled_dot_product_attention_out)
   layer_norm_1 = LayerNormalization <axis: int = -1, epsilon: float = 1e-06, stash_type: int = 1> (add_119, "text.transformer.resblocks.0.ln_2.weight", "text.transformer.resblocks.0.ln_2.bias")
   val_41 = MatMul (layer_norm_1, val_5)
   linear_2 = Add (val_41, "text.transformer.resblocks.0.mlp.c_fc.bias")
   [node_gelu] gelu = Gelu <approximate: string = "none"> (linear_2)
   val_42 = MatMul (gelu, val_6)
   linear_3 = Add (val_42, "text.transformer.resblocks.0.mlp.c_proj.bias")
   add_140 = Add (add_119, linear_3)
   layer_norm_2 = LayerNormalization <axis: int = -1, epsilon: float = 1e-06, stash_type: int = 1> (add_140, "text.transformer.resblocks.1.ln_1.weight", "text.transformer.resblocks.1.ln_1.bias")
   [node_scaled_dot_product_attention_1_qkv_mm] node_scaled_dot_product_attention_1_qkv_mm_out = MatMul (layer_norm_2, val_7)
   [node_scaled_dot_product_attention_1_qkv_bias] node_scaled_dot_product_attention_1_qkv = Add (node_scaled_dot_product_attention_1_qkv_mm_out, "text.transformer.resblocks.1.attn.in_proj_bias")
   [node_scaled_dot_product_attention_1_qkv_split] node_scaled_dot_product_attention_1_q, node_scaled_dot_product_attention_1_k, node_scaled_dot_product_attention_1_v = Split <axis: int = -1> (node_scaled_dot_product_attention_1_qkv, attn3d_split_3x768)
   scaled_dot_product_attention_1 = Attention <is_causal: int = 0, kv_num_heads: int = 12, q_num_heads: int = 12, qk_matmul_output_mode: int = 0, softcap: float = 0> (node_scaled_dot_product_attention_1_q, node_scaled_dot_product_attention_1_k, node_scaled_dot_product_attention_1_v)
   [node_scaled_dot_product_attention_1_out_mm] node_scaled_dot_product_attention_1_out_mm_out = MatMul (scaled_dot_product_attention_1, node_scaled_dot_product_attention_1_wo_t)
   [node_scaled_dot_product_attention_1_out_bias] node_scaled_dot_product_attention_1_out = Add (node_scaled_dot_product_attention_1_out_mm_out, "text.transformer.resblocks.1.attn.out_proj.bias")
   add_255 = Add (add_140, node_scaled_dot_product_attention_1_out)
   layer_norm_3 = LayerNormalization <axis: int = -1, epsilon: float = 1e-06, stash_type: int = 1> (add_255, "text.transformer.resblocks.1.ln_2.weight", "text.transformer.resblocks.1.ln_2.bias")
   val_43 = MatMul (layer_norm_3, val_8)
   linear_6 = Add (val_43, "text.transformer.resblocks.1.mlp.c_fc.bias")
   gelu_1 = Gelu <approximate: string = "none"> (linear_6)
   val_44 = MatMul (gelu_1, val_9)
   linear_7 = Add (val_44, "text.transformer.resblocks.1.mlp.c_proj.bias")
   add_276 = Add (add_255, linear_7)
   layer_norm_4 = LayerNormalization <axis: int = -1, epsilon: float = 1e-06, stash_type: int = 1> (add_276, "text.transformer.resblocks.2.ln_1.weight", "text.transformer.resblocks.2.ln_1.bias")
   [node_scaled_dot_product_attention_2_qkv_mm] node_scaled_dot_product_attention_2_qkv_mm_out = MatMul (layer_norm_4, val_10)
   [node_scaled_dot_product_attention_2_qkv_bias] node_scaled_dot_product_attention_2_qkv = Add (node_scaled_dot_product_attention_2_qkv_mm_out, "text.transformer.resblocks.2.attn.in_proj_bias")
   [node_scaled_dot_product_attention_2_qkv_split] node_scaled_dot_product_attention_2_q, node_scaled_dot_product_attention_2_k, node_scaled_dot_product_attention_2_v = Split <axis: int = -1> (node_scaled_dot_product_attention_2_qkv, attn3d_split_3x768)
   scaled_dot_product_attention_2 = Attention <is_causal: int = 0, kv_num_heads: int = 12, q_num_heads: int = 12, qk_matmul_output_mode: int = 0, softcap: float = 0> (node_scaled_dot_product_attention_2_q, node_scaled_dot_product_attention_2_k, node_scaled_dot_product_attention_2_v)
   [node_scaled_dot_product_attention_2_out_mm] node_scaled_dot_product_attention_2_out_mm_out = MatMul (scaled_dot_product_attention_2, node_scaled_dot_product_attention_2_wo_t)
   [node_scaled_dot_product_attention_2_out_bias] node_scaled_dot_product_attention_2_out = Add (node_scaled_dot_product_attention_2_out_mm_out, "text.transformer.resblocks.2.attn.out_proj.bias")
   add_391 = Add (add_276, node_scaled_dot_product_attention_2_out)
   layer_norm_5 = LayerNormalization <axis: int = -1, epsilon: float = 1e-06, stash_type: int = 1> (add_391, "text.transformer.resblocks.2.ln_2.weight", "text.transformer.resblocks.2.ln_2.bias")
   val_45 = MatMul (layer_norm_5, val_11)
   linear_10 = Add (val_45, "text.transformer.resblocks.2.mlp.c_fc.bias")
   gelu_2 = Gelu <approximate: string = "none"> (linear_10)
   val_46 = MatMul (gelu_2, val_12)
   linear_11 = Add (val_46, "text.transformer.resblocks.2.mlp.c_proj.bias")
   add_412 = Add (add_391, linear_11)
   layer_norm_6 = LayerNormalization <axis: int = -1, epsilon: float = 1e-06, stash_type: int = 1> (add_412, "text.transformer.resblocks.3.ln_1.weight", "text.transformer.resblocks.3.ln_1.bias")
   [node_scaled_dot_product_attention_3_qkv_mm] node_scaled_dot_product_attention_3_qkv_mm_out = MatMul (layer_norm_6, val_13)
   [node_scaled_dot_product_attention_3_qkv_bias] node_scaled_dot_product_attention_3_qkv = Add (node_scaled_dot_product_attention_3_qkv_mm_out, "text.transformer.resblocks.3.attn.in_proj_bias")
   [node_scaled_dot_product_attention_3_qkv_split] node_scaled_dot_product_attention_3_q, node_scaled_dot_product_attention_3_k, node_scaled_dot_product_attention_3_v = Split <axis: int = -1> (node_scaled_dot_product_attention_3_qkv, attn3d_split_3x768)
   scaled_dot_product_attention_3 = Attention <is_causal: int = 0, kv_num_heads: int = 12, q_num_heads: int = 12, qk_matmul_output_mode: int = 0, softcap: float = 0> (node_scaled_dot_product_attention_3_q, node_scaled_dot_product_attention_3_k, node_scaled_dot_product_attention_3_v)
   [node_scaled_dot_product_attention_3_out_mm] node_scaled_dot_product_attention_3_out_mm_out = MatMul (scaled_dot_product_attention_3, node_scaled_dot_product_attention_3_wo_t)
   [node_scaled_dot_product_attention_3_out_bias] node_scaled_dot_product_attention_3_out = Add (node_scaled_dot_product_attention_3_out_mm_out, "text.transformer.resblocks.3.attn.out_proj.bias")
   add_527 = Add (add_412, node_scaled_dot_product_attention_3_out)
   layer_norm_7 = LayerNormalization <axis: int = -1, epsilon: float = 1e-06, stash_type: int = 1> (add_527, "text.transformer.resblocks.3.ln_2.weight", "text.transformer.resblocks.3.ln_2.bias")
   val_47 = MatMul (layer_norm_7, val_14)
   linear_14 = Add (val_47, "text.transformer.resblocks.3.mlp.c_fc.bias")
   gelu_3 = Gelu <approximate: string = "none"> (linear_14)
   val_48 = MatMul (gelu_3, val_15)
   linear_15 = Add (val_48, "text.transformer.resblocks.3.mlp.c_proj.bias")
   add_548 = Add (add_527, linear_15)
   layer_norm_8 = LayerNormalization <axis: int = -1, epsilon: float = 1e-06, stash_type: int = 1> (add_548, "text.transformer.resblocks.4.ln_1.weight", "text.transformer.resblocks.4.ln_1.bias")
   [node_scaled_dot_product_attention_4_qkv_mm] node_scaled_dot_product_attention_4_qkv_mm_out = MatMul (layer_norm_8, val_16)
   [node_scaled_dot_product_attention_4_qkv_bias] node_scaled_dot_product_attention_4_qkv = Add (node_scaled_dot_product_attention_4_qkv_mm_out, "text.transformer.resblocks.4.attn.in_proj_bias")
   [node_scaled_dot_product_attention_4_qkv_split] node_scaled_dot_product_attention_4_q, node_scaled_dot_product_attention_4_k, node_scaled_dot_product_attention_4_v = Split <axis: int = -1> (node_scaled_dot_product_attention_4_qkv, attn3d_split_3x768)
   scaled_dot_product_attention_4 = Attention <is_causal: int = 0, kv_num_heads: int = 12, q_num_heads: int = 12, qk_matmul_output_mode: int = 0, softcap: float = 0> (node_scaled_dot_product_attention_4_q, node_scaled_dot_product_attention_4_k, node_scaled_dot_product_attention_4_v)
   [node_scaled_dot_product_attention_4_out_mm] node_scaled_dot_product_attention_4_out_mm_out = MatMul (scaled_dot_product_attention_4, node_scaled_dot_product_attention_4_wo_t)
   [node_scaled_dot_product_attention_4_out_bias] node_scaled_dot_product_attention_4_out = Add (node_scaled_dot_product_attention_4_out_mm_out, "text.transformer.resblocks.4.attn.out_proj.bias")
   add_663 = Add (add_548, node_scaled_dot_product_attention_4_out)
   layer_norm_9 = LayerNormalization <axis: int = -1, epsilon: float = 1e-06, stash_type: int = 1> (add_663, "text.transformer.resblocks.4.ln_2.weight", "text.transformer.resblocks.4.ln_2.bias")
   val_49 = MatMul (layer_norm_9, val_17)
   linear_18 = Add (val_49, "text.transformer.resblocks.4.mlp.c_fc.bias")
   gelu_4 = Gelu <approximate: string = "none"> (linear_18)
   val_50 = MatMul (gelu_4, val_18)
   linear_19 = Add (val_50, "text.transformer.resblocks.4.mlp.c_proj.bias")
   add_684 = Add (add_663, linear_19)
   layer_norm_10 = LayerNormalization <axis: int = -1, epsilon: float = 1e-06, stash_type: int = 1> (add_684, "text.transformer.resblocks.5.ln_1.weight", "text.transformer.resblocks.5.ln_1.bias")
   [node_scaled_dot_product_attention_5_qkv_mm] node_scaled_dot_product_attention_5_qkv_mm_out = MatMul (layer_norm_10, val_19)
   [node_scaled_dot_product_attention_5_qkv_bias] node_scaled_dot_product_attention_5_qkv = Add (node_scaled_dot_product_attention_5_qkv_mm_out, "text.transformer.resblocks.5.attn.in_proj_bias")
   [node_scaled_dot_product_attention_5_qkv_split] node_scaled_dot_product_attention_5_q, node_scaled_dot_product_attention_5_k, node_scaled_dot_product_attention_5_v = Split <axis: int = -1> (node_scaled_dot_product_attention_5_qkv, attn3d_split_3x768)
   scaled_dot_product_attention_5 = Attention <is_causal: int = 0, kv_num_heads: int = 12, q_num_heads: int = 12, qk_matmul_output_mode: int = 0, softcap: float = 0> (node_scaled_dot_product_attention_5_q, node_scaled_dot_product_attention_5_k, node_scaled_dot_product_attention_5_v)
   [node_scaled_dot_product_attention_5_out_mm] node_scaled_dot_product_attention_5_out_mm_out = MatMul (scaled_dot_product_attention_5, node_scaled_dot_product_attention_5_wo_t)
   [node_scaled_dot_product_attention_5_out_bias] node_scaled_dot_product_attention_5_out = Add (node_scaled_dot_product_attention_5_out_mm_out, "text.transformer.resblocks.5.attn.out_proj.bias")
   add_799 = Add (add_684, node_scaled_dot_product_attention_5_out)
   layer_norm_11 = LayerNormalization <axis: int = -1, epsilon: float = 1e-06, stash_type: int = 1> (add_799, "text.transformer.resblocks.5.ln_2.weight", "text.transformer.resblocks.5.ln_2.bias")
   val_51 = MatMul (layer_norm_11, val_20)
   linear_22 = Add (val_51, "text.transformer.resblocks.5.mlp.c_fc.bias")
   gelu_5 = Gelu <approximate: string = "none"> (linear_22)
   val_52 = MatMul (gelu_5, val_21)
   linear_23 = Add (val_52, "text.transformer.resblocks.5.mlp.c_proj.bias")
   add_820 = Add (add_799, linear_23)
   layer_norm_12 = LayerNormalization <axis: int = -1, epsilon: float = 1e-06, stash_type: int = 1> (add_820, "text.transformer.resblocks.6.ln_1.weight", "text.transformer.resblocks.6.ln_1.bias")
   [node_scaled_dot_product_attention_6_qkv_mm] node_scaled_dot_product_attention_6_qkv_mm_out = MatMul (layer_norm_12, val_22)
   [node_scaled_dot_product_attention_6_qkv_bias] node_scaled_dot_product_attention_6_qkv = Add (node_scaled_dot_product_attention_6_qkv_mm_out, "text.transformer.resblocks.6.attn.in_proj_bias")
   [node_scaled_dot_product_attention_6_qkv_split] node_scaled_dot_product_attention_6_q, node_scaled_dot_product_attention_6_k, node_scaled_dot_product_attention_6_v = Split <axis: int = -1> (node_scaled_dot_product_attention_6_qkv, attn3d_split_3x768)
   scaled_dot_product_attention_6 = Attention <is_causal: int = 0, kv_num_heads: int = 12, q_num_heads: int = 12, qk_matmul_output_mode: int = 0, softcap: float = 0> (node_scaled_dot_product_attention_6_q, node_scaled_dot_product_attention_6_k, node_scaled_dot_product_attention_6_v)
   [node_scaled_dot_product_attention_6_out_mm] node_scaled_dot_product_attention_6_out_mm_out = MatMul (scaled_dot_product_attention_6, node_scaled_dot_product_attention_6_wo_t)
   [node_scaled_dot_product_attention_6_out_bias] node_scaled_dot_product_attention_6_out = Add (node_scaled_dot_product_attention_6_out_mm_out, "text.transformer.resblocks.6.attn.out_proj.bias")
   add_935 = Add (add_820, node_scaled_dot_product_attention_6_out)
   layer_norm_13 = LayerNormalization <axis: int = -1, epsilon: float = 1e-06, stash_type: int = 1> (add_935, "text.transformer.resblocks.6.ln_2.weight", "text.transformer.resblocks.6.ln_2.bias")
   val_53 = MatMul (layer_norm_13, val_23)
   linear_26 = Add (val_53, "text.transformer.resblocks.6.mlp.c_fc.bias")
   gelu_6 = Gelu <approximate: string = "none"> (linear_26)
   val_54 = MatMul (gelu_6, val_24)
   linear_27 = Add (val_54, "text.transformer.resblocks.6.mlp.c_proj.bias")
   add_956 = Add (add_935, linear_27)
   layer_norm_14 = LayerNormalization <axis: int = -1, epsilon: float = 1e-06, stash_type: int = 1> (add_956, "text.transformer.resblocks.7.ln_1.weight", "text.transformer.resblocks.7.ln_1.bias")
   [node_scaled_dot_product_attention_7_qkv_mm] node_scaled_dot_product_attention_7_qkv_mm_out = MatMul (layer_norm_14, val_25)
   [node_scaled_dot_product_attention_7_qkv_bias] node_scaled_dot_product_attention_7_qkv = Add (node_scaled_dot_product_attention_7_qkv_mm_out, "text.transformer.resblocks.7.attn.in_proj_bias")
   [node_scaled_dot_product_attention_7_qkv_split] node_scaled_dot_product_attention_7_q, node_scaled_dot_product_attention_7_k, node_scaled_dot_product_attention_7_v = Split <axis: int = -1> (node_scaled_dot_product_attention_7_qkv, attn3d_split_3x768)
   scaled_dot_product_attention_7 = Attention <is_causal: int = 0, kv_num_heads: int = 12, q_num_heads: int = 12, qk_matmul_output_mode: int = 0, softcap: float = 0> (node_scaled_dot_product_attention_7_q, node_scaled_dot_product_attention_7_k, node_scaled_dot_product_attention_7_v)
   [node_scaled_dot_product_attention_7_out_mm] node_scaled_dot_product_attention_7_out_mm_out = MatMul (scaled_dot_product_attention_7, node_scaled_dot_product_attention_7_wo_t)
   [node_scaled_dot_product_attention_7_out_bias] node_scaled_dot_product_attention_7_out = Add (node_scaled_dot_product_attention_7_out_mm_out, "text.transformer.resblocks.7.attn.out_proj.bias")
   add_1071 = Add (add_956, node_scaled_dot_product_attention_7_out)
   layer_norm_15 = LayerNormalization <axis: int = -1, epsilon: float = 1e-06, stash_type: int = 1> (add_1071, "text.transformer.resblocks.7.ln_2.weight", "text.transformer.resblocks.7.ln_2.bias")
   val_55 = MatMul (layer_norm_15, val_26)
   linear_30 = Add (val_55, "text.transformer.resblocks.7.mlp.c_fc.bias")
   gelu_7 = Gelu <approximate: string = "none"> (linear_30)
   val_56 = MatMul (gelu_7, val_27)
   linear_31 = Add (val_56, "text.transformer.resblocks.7.mlp.c_proj.bias")
   add_1092 = Add (add_1071, linear_31)
   layer_norm_16 = LayerNormalization <axis: int = -1, epsilon: float = 1e-06, stash_type: int = 1> (add_1092, "text.transformer.resblocks.8.ln_1.weight", "text.transformer.resblocks.8.ln_1.bias")
   [node_scaled_dot_product_attention_8_qkv_mm] node_scaled_dot_product_attention_8_qkv_mm_out = MatMul (layer_norm_16, val_28)
   [node_scaled_dot_product_attention_8_qkv_bias] node_scaled_dot_product_attention_8_qkv = Add (node_scaled_dot_product_attention_8_qkv_mm_out, "text.transformer.resblocks.8.attn.in_proj_bias")
   [node_scaled_dot_product_attention_8_qkv_split] node_scaled_dot_product_attention_8_q, node_scaled_dot_product_attention_8_k, node_scaled_dot_product_attention_8_v = Split <axis: int = -1> (node_scaled_dot_product_attention_8_qkv, attn3d_split_3x768)
   scaled_dot_product_attention_8 = Attention <is_causal: int = 0, kv_num_heads: int = 12, q_num_heads: int = 12, qk_matmul_output_mode: int = 0, softcap: float = 0> (node_scaled_dot_product_attention_8_q, node_scaled_dot_product_attention_8_k, node_scaled_dot_product_attention_8_v)
   [node_scaled_dot_product_attention_8_out_mm] node_scaled_dot_product_attention_8_out_mm_out = MatMul (scaled_dot_product_attention_8, node_scaled_dot_product_attention_8_wo_t)
   [node_scaled_dot_product_attention_8_out_bias] node_scaled_dot_product_attention_8_out = Add (node_scaled_dot_product_attention_8_out_mm_out, "text.transformer.resblocks.8.attn.out_proj.bias")
   add_1207 = Add (add_1092, node_scaled_dot_product_attention_8_out)
   layer_norm_17 = LayerNormalization <axis: int = -1, epsilon: float = 1e-06, stash_type: int = 1> (add_1207, "text.transformer.resblocks.8.ln_2.weight", "text.transformer.resblocks.8.ln_2.bias")
   val_57 = MatMul (layer_norm_17, val_29)
   linear_34 = Add (val_57, "text.transformer.resblocks.8.mlp.c_fc.bias")
   gelu_8 = Gelu <approximate: string = "none"> (linear_34)
   val_58 = MatMul (gelu_8, val_30)
   linear_35 = Add (val_58, "text.transformer.resblocks.8.mlp.c_proj.bias")
   add_1228 = Add (add_1207, linear_35)
   layer_norm_18 = LayerNormalization <axis: int = -1, epsilon: float = 1e-06, stash_type: int = 1> (add_1228, "text.transformer.resblocks.9.ln_1.weight", "text.transformer.resblocks.9.ln_1.bias")
   [node_scaled_dot_product_attention_9_qkv_mm] node_scaled_dot_product_attention_9_qkv_mm_out = MatMul (layer_norm_18, val_31)
   [node_scaled_dot_product_attention_9_qkv_bias] node_scaled_dot_product_attention_9_qkv = Add (node_scaled_dot_product_attention_9_qkv_mm_out, "text.transformer.resblocks.9.attn.in_proj_bias")
   [node_scaled_dot_product_attention_9_qkv_split] node_scaled_dot_product_attention_9_q, node_scaled_dot_product_attention_9_k, node_scaled_dot_product_attention_9_v = Split <axis: int = -1> (node_scaled_dot_product_attention_9_qkv, attn3d_split_3x768)
   scaled_dot_product_attention_9 = Attention <is_causal: int = 0, kv_num_heads: int = 12, q_num_heads: int = 12, qk_matmul_output_mode: int = 0, softcap: float = 0> (node_scaled_dot_product_attention_9_q, node_scaled_dot_product_attention_9_k, node_scaled_dot_product_attention_9_v)
   [node_scaled_dot_product_attention_9_out_mm] node_scaled_dot_product_attention_9_out_mm_out = MatMul (scaled_dot_product_attention_9, node_scaled_dot_product_attention_9_wo_t)
   [node_scaled_dot_product_attention_9_out_bias] node_scaled_dot_product_attention_9_out = Add (node_scaled_dot_product_attention_9_out_mm_out, "text.transformer.resblocks.9.attn.out_proj.bias")
   add_1343 = Add (add_1228, node_scaled_dot_product_attention_9_out)
   layer_norm_19 = LayerNormalization <axis: int = -1, epsilon: float = 1e-06, stash_type: int = 1> (add_1343, "text.transformer.resblocks.9.ln_2.weight", "text.transformer.resblocks.9.ln_2.bias")
   val_59 = MatMul (layer_norm_19, val_32)
   linear_38 = Add (val_59, "text.transformer.resblocks.9.mlp.c_fc.bias")
   gelu_9 = Gelu <approximate: string = "none"> (linear_38)
   val_60 = MatMul (gelu_9, val_33)
   linear_39 = Add (val_60, "text.transformer.resblocks.9.mlp.c_proj.bias")
   add_1364 = Add (add_1343, linear_39)
   layer_norm_20 = LayerNormalization <axis: int = -1, epsilon: float = 1e-06, stash_type: int = 1> (add_1364, "text.transformer.resblocks.10.ln_1.weight", "text.transformer.resblocks.10.ln_1.bias")
   [node_scaled_dot_product_attention_10_qkv_mm] node_scaled_dot_product_attention_10_qkv_mm_out = MatMul (layer_norm_20, val_34)
   [node_scaled_dot_product_attention_10_qkv_bias] node_scaled_dot_product_attention_10_qkv = Add (node_scaled_dot_product_attention_10_qkv_mm_out, "text.transformer.resblocks.10.attn.in_proj_bias")
   [node_scaled_dot_product_attention_10_qkv_split] node_scaled_dot_product_attention_10_q, node_scaled_dot_product_attention_10_k, node_scaled_dot_product_attention_10_v = Split <axis: int = -1> (node_scaled_dot_product_attention_10_qkv, attn3d_split_3x768)
   scaled_dot_product_attention_10 = Attention <is_causal: int = 0, kv_num_heads: int = 12, q_num_heads: int = 12, qk_matmul_output_mode: int = 0, softcap: float = 0> (node_scaled_dot_product_attention_10_q, node_scaled_dot_product_attention_10_k, node_scaled_dot_product_attention_10_v)
   [node_scaled_dot_product_attention_10_out_mm] node_scaled_dot_product_attention_10_out_mm_out = MatMul (scaled_dot_product_attention_10, node_scaled_dot_product_attention_10_wo_t)
   [node_scaled_dot_product_attention_10_out_bias] node_scaled_dot_product_attention_10_out = Add (node_scaled_dot_product_attention_10_out_mm_out, "text.transformer.resblocks.10.attn.out_proj.bias")
   add_1479 = Add (add_1364, node_scaled_dot_product_attention_10_out)
   layer_norm_21 = LayerNormalization <axis: int = -1, epsilon: float = 1e-06, stash_type: int = 1> (add_1479, "text.transformer.resblocks.10.ln_2.weight", "text.transformer.resblocks.10.ln_2.bias")
   val_61 = MatMul (layer_norm_21, val_35)
   linear_42 = Add (val_61, "text.transformer.resblocks.10.mlp.c_fc.bias")
   gelu_10 = Gelu <approximate: string = "none"> (linear_42)
   val_62 = MatMul (gelu_10, val_36)
   linear_43 = Add (val_62, "text.transformer.resblocks.10.mlp.c_proj.bias")
   add_1500 = Add (add_1479, linear_43)
   layer_norm_22 = LayerNormalization <axis: int = -1, epsilon: float = 1e-06, stash_type: int = 1> (add_1500, "text.transformer.resblocks.11.ln_1.weight", "text.transformer.resblocks.11.ln_1.bias")
   [node_scaled_dot_product_attention_11_qkv_mm] node_scaled_dot_product_attention_11_qkv_mm_out = MatMul (layer_norm_22, val_37)
   [node_scaled_dot_product_attention_11_qkv_bias] node_scaled_dot_product_attention_11_qkv = Add (node_scaled_dot_product_attention_11_qkv_mm_out, "text.transformer.resblocks.11.attn.in_proj_bias")
   [node_scaled_dot_product_attention_11_qkv_split] node_scaled_dot_product_attention_11_q, node_scaled_dot_product_attention_11_k, node_scaled_dot_product_attention_11_v = Split <axis: int = -1> (node_scaled_dot_product_attention_11_qkv, attn3d_split_3x768)
   [pool_hoist_node_scaled_dot_product_attention_11_q] node_scaled_dot_product_attention_11_q_pooled = Slice (node_scaled_dot_product_attention_11_q, val_0, val_3, val_2)
   scaled_dot_product_attention_11 = Attention <is_causal: int = 0, kv_num_heads: int = 12, q_num_heads: int = 12, qk_matmul_output_mode: int = 0, softcap: float = 0> (node_scaled_dot_product_attention_11_q_pooled, node_scaled_dot_product_attention_11_k, node_scaled_dot_product_attention_11_v)
   [node_scaled_dot_product_attention_11_out_mm] node_scaled_dot_product_attention_11_out_mm_out = MatMul (scaled_dot_product_attention_11, node_scaled_dot_product_attention_11_wo_t)
   [node_scaled_dot_product_attention_11_out_bias] node_scaled_dot_product_attention_11_out = Add (node_scaled_dot_product_attention_11_out_mm_out, "text.transformer.resblocks.11.attn.out_proj.bias")
   [pool_hoist_add_1500] add_1500_pooled = Slice (add_1500, val_0, val_3, val_2)
   add_1615 = Add (add_1500_pooled, node_scaled_dot_product_attention_11_out)
   layer_norm_23 = LayerNormalization <axis: int = -1, epsilon: float = 1e-06, stash_type: int = 1> (add_1615, "text.transformer.resblocks.11.ln_2.weight", "text.transformer.resblocks.11.ln_2.bias")
   val_63 = MatMul (layer_norm_23, val_38)
   linear_46 = Add (val_63, "text.transformer.resblocks.11.mlp.c_fc.bias")
   gelu_11 = Gelu <approximate: string = "none"> (linear_46)
   val_64 = MatMul (gelu_11, val_39)
   linear_47 = Add (val_64, "text.transformer.resblocks.11.mlp.c_proj.bias")
   add_1636 = Add (add_1615, linear_47)
   val_65 = Squeeze (add_1636, val_2)
   select_36 = LayerNormalization <axis: int = -1, epsilon: float = 1e-06, stash_type: int = 1> (val_65, "text.ln_final.weight", "text.ln_final.bias")
   linear_48 = Gemm <alpha: float = 1, beta: float = 1, transA: int = 0, transB: int = 1> (select_36, "text.text_projection.weight", "text.text_projection.bias")
   [node_linalg_vector_norm] linalg_vector_norm = ReduceL2 <keepdims: int = 1, noop_with_empty_axes: int = 0> (linear_48, val_0)
   [node_clamp_min] clamp_min = Clip (linalg_vector_norm, val_1)
   [node_div] text_embedding = Div (linear_48, clamp_min)
}

weights:
attn3d_split_3x768 INT64[3] eca50b2daf34
node_scaled_dot_product_attention_10_wo_t FLOAT[768,768] 99d641c43921
node_scaled_dot_product_attention_11_wo_t FLOAT[768,768] f63c4907cf25
node_scaled_dot_product_attention_1_wo_t FLOAT[768,768] 2e7a8a9daa18
node_scaled_dot_product_attention_2_wo_t FLOAT[768,768] 4665233ce048
node_scaled_dot_product_attention_3_wo_t FLOAT[768,768] 2b062970103e
node_scaled_dot_product_attention_4_wo_t FLOAT[768,768] 94fefabc678f
node_scaled_dot_product_attention_5_wo_t FLOAT[768,768] 7d08ec807ad2
node_scaled_dot_product_attention_6_wo_t FLOAT[768,768] 6a3fa7f2d345
node_scaled_dot_product_attention_7_wo_t FLOAT[768,768] ee6684d21f4f
node_scaled_dot_product_attention_8_wo_t FLOAT[768,768] 5b252a2abf1d
node_scaled_dot_product_attention_9_wo_t FLOAT[768,768] 16c5431008fc
node_scaled_dot_product_attention_wo_t FLOAT[768,768] 5fbc1d7280d3
text.ln_final.bias FLOAT[768] c34a98b95a68
text.ln_final.weight FLOAT[768] b3ac9475265e
text.positional_embedding FLOAT[64,768] 2a9ff11479b7
text.text_projection.bias FLOAT[768] 89aea87697af
text.text_projection.weight FLOAT[768,768] edbdc89b11b8
text.token_embedding.weight_fp16 FLOAT16[250000,768] 374ea2c0e2f0
text.transformer.resblocks.0.attn.in_proj_bias FLOAT[2304] 1d63ff75f4cb
text.transformer.resblocks.0.attn.out_proj.bias FLOAT[768] 1a772042ed81
text.transformer.resblocks.0.ln_1.bias FLOAT[768] e80788a1492f
text.transformer.resblocks.0.ln_1.weight FLOAT[768] dc0187141516
text.transformer.resblocks.0.ln_2.bias FLOAT[768] eb935be1409c
text.transformer.resblocks.0.ln_2.weight FLOAT[768] 07ce61e2a610
text.transformer.resblocks.0.mlp.c_fc.bias FLOAT[3072] 1e0373c3bfc1
text.transformer.resblocks.0.mlp.c_proj.bias FLOAT[768] f53aef54690c
text.transformer.resblocks.1.attn.in_proj_bias FLOAT[2304] 0090889e87fc
text.transformer.resblocks.1.attn.out_proj.bias FLOAT[768] b2fe377badb7
text.transformer.resblocks.1.ln_1.bias FLOAT[768] 2eb82ef8df2e
text.transformer.resblocks.1.ln_1.weight FLOAT[768] e6e3a90268b1
text.transformer.resblocks.1.ln_2.bias FLOAT[768] 62c8537bbbd6
text.transformer.resblocks.1.ln_2.weight FLOAT[768] fa5e9a39e253
text.transformer.resblocks.1.mlp.c_fc.bias FLOAT[3072] 1a4696696dda
text.transformer.resblocks.1.mlp.c_proj.bias FLOAT[768] 9e63c064887c
text.transformer.resblocks.10.attn.in_proj_bias FLOAT[2304] 19cdf3ed088b
text.transformer.resblocks.10.attn.out_proj.bias FLOAT[768] e334c099159c
text.transformer.resblocks.10.ln_1.bias FLOAT[768] 67f007d3f632
text.transformer.resblocks.10.ln_1.weight FLOAT[768] 9e53c8e9a884
text.transformer.resblocks.10.ln_2.bias FLOAT[768] e7428524b783
text.transformer.resblocks.10.ln_2.weight FLOAT[768] 5001d48cdd37
text.transformer.resblocks.10.mlp.c_fc.bias FLOAT[3072] 7e757501f921
text.transformer.resblocks.10.mlp.c_proj.bias FLOAT[768] 6d22e40dbd8a
text.transformer.resblocks.11.attn.in_proj_bias FLOAT[2304] f3e77b3887ad
text.transformer.resblocks.11.attn.out_proj.bias FLOAT[768] 267d0825b959
text.transformer.resblocks.11.ln_1.bias FLOAT[768] 4512a1cb20cf
text.transformer.resblocks.11.ln_1.weight FLOAT[768] 0415fc5da59c
text.transformer.resblocks.11.ln_2.bias FLOAT[768] 9424661c79e2
text.transformer.resblocks.11.ln_2.weight FLOAT[768] 81504b43352d
text.transformer.resblocks.11.mlp.c_fc.bias FLOAT[3072] 1fc24f5fa3b6
text.transformer.resblocks.11.mlp.c_proj.bias FLOAT[768] a3cef3ca4b83
text.transformer.resblocks.2.attn.in_proj_bias FLOAT[2304] 61d0da43c323
text.transformer.resblocks.2.attn.out_proj.bias FLOAT[768] f5b83b02c5e7
text.transformer.resblocks.2.ln_1.bias FLOAT[768] 96c4e143b649
text.transformer.resblocks.2.ln_1.weight FLOAT[768] c3af1333ea4b
text.transformer.resblocks.2.ln_2.bias FLOAT[768] 9de286180adb
text.transformer.resblocks.2.ln_2.weight FLOAT[768] 6e72b5947969
text.transformer.resblocks.2.mlp.c_fc.bias FLOAT[3072] f5a5c9ceb772
text.transformer.resblocks.2.mlp.c_proj.bias FLOAT[768] 31d4ea9b95ea
text.transformer.resblocks.3.attn.in_proj_bias FLOAT[2304] 88fdb45446d5
text.transformer.resblocks.3.attn.out_proj.bias FLOAT[768] de5d5d3080d5
text.transformer.resblocks.3.ln_1.bias FLOAT[768] 8183d3f91410
text.transformer.resblocks.3.ln_1.weight FLOAT[768] 78c2193abf68
text.transformer.resblocks.3.ln_2.bias FLOAT[768] 699b66e1c4dd
text.transformer.resblocks.3.ln_2.weight FLOAT[768] efd3ca1c09c7
text.transformer.resblocks.3.mlp.c_fc.bias FLOAT[3072] 4a7c1a6ab6a5
text.transformer.resblocks.3.mlp.c_proj.bias FLOAT[768] b545b9e5b324
text.transformer.resblocks.4.attn.in_proj_bias FLOAT[2304] 42b7c90054eb
text.transformer.resblocks.4.attn.out_proj.bias FLOAT[768] bd4cb705d15d
text.transformer.resblocks.4.ln_1.bias FLOAT[768] 22b024c60076
text.transformer.resblocks.4.ln_1.weight FLOAT[768] aa18110c0f85
text.transformer.resblocks.4.ln_2.bias FLOAT[768] 8d7b78a66759
text.transformer.resblocks.4.ln_2.weight FLOAT[768] 6f9136a6613c
text.transformer.resblocks.4.mlp.c_fc.bias FLOAT[3072] 5470c83680d9
text.transformer.resblocks.4.mlp.c_proj.bias FLOAT[768] da59adb870e8
text.transformer.resblocks.5.attn.in_proj_bias FLOAT[2304] 7adc98d68c8b
text.transformer.resblocks.5.attn.out_proj.bias FLOAT[768] 1f9ab3e07b4d
text.transformer.resblocks.5.ln_1.bias FLOAT[768] 5f59b748457d
text.transformer.resblocks.5.ln_1.weight FLOAT[768] 4f42b2368354
text.transformer.resblocks.5.ln_2.bias FLOAT[768] 6a5d8fd1f57c
text.transformer.resblocks.5.ln_2.weight FLOAT[768] 41ef4560332e
text.transformer.resblocks.5.mlp.c_fc.bias FLOAT[3072] 6ea843d06d6f
text.transformer.resblocks.5.mlp.c_proj.bias FLOAT[768] 3ba09abc72e4
text.transformer.resblocks.6.attn.in_proj_bias FLOAT[2304] 1a9b5b673c8f
text.transformer.resblocks.6.attn.out_proj.bias FLOAT[768] 895d3f9d24e0
text.transformer.resblocks.6.ln_1.bias FLOAT[768] aa96239e5013
text.transformer.resblocks.6.ln_1.weight FLOAT[768] 75b9a14b1927
text.transformer.resblocks.6.ln_2.bias FLOAT[768] ac7a29794b0f
text.transformer.resblocks.6.ln_2.weight FLOAT[768] b64fa2da3452
text.transformer.resblocks.6.mlp.c_fc.bias FLOAT[3072] 87148e21d8cf
text.transformer.resblocks.6.mlp.c_proj.bias FLOAT[768] 09b5a2d09824
text.transformer.resblocks.7.attn.in_proj_bias FLOAT[2304] 2aa7ee25a9c0
text.transformer.resblocks.7.attn.out_proj.bias FLOAT[768] a37865e3deda
text.transformer.resblocks.7.ln_1.bias FLOAT[768] 3787e03dabc4
text.transformer.resblocks.7.ln_1.weight FLOAT[768] 9545528674e4
text.transformer.resblocks.7.ln_2.bias FLOAT[768] 44e4edfa46b8
text.transformer.resblocks.7.ln_2.weight FLOAT[768] 514e041120af
text.transformer.resblocks.7.mlp.c_fc.bias FLOAT[3072] 10b4e283af81
text.transformer.resblocks.7.mlp.c_proj.bias FLOAT[768] 301db5f3e50d
text.transformer.resblocks.8.attn.in_proj_bias FLOAT[2304] 0605c31d11d4
text.transformer.resblocks.8.attn.out_proj.bias FLOAT[768] 44a74dc5cfc9
text.transformer.resblocks.8.ln_1.bias FLOAT[768] 3512d3d2474e
text.transformer.resblocks.8.ln_1.weight FLOAT[768] 3ba01671177c
text.transformer.resblocks.8.ln_2.bias FLOAT[768] c2c8544eea1b
text.transformer.resblocks.8.ln_2.weight FLOAT[768] c554a2659ead
text.transformer.resblocks.8.mlp.c_fc.bias FLOAT[3072] 33e3792a2525
text.transformer.resblocks.8.mlp.c_proj.bias FLOAT[768] 8a934076786b
text.transformer.resblocks.9.attn.in_proj_bias FLOAT[2304] a61e175a7e49
text.transformer.resblocks.9.attn.out_proj.bias FLOAT[768] d9051c811b46
text.transformer.resblocks.9.ln_1.bias FLOAT[768] 596aa7812eb4
text.transformer.resblocks.9.ln_1.weight FLOAT[768] 049a09a1840d
text.transformer.resblocks.9.ln_2.bias FLOAT[768] 1f2d8790c582
text.transformer.resblocks.9.ln_2.weight FLOAT[768] 3c38f00292d8
text.transformer.resblocks.9.mlp.c_fc.bias FLOAT[3072] 4f1d07255c03
text.transformer.resblocks.9.mlp.c_proj.bias FLOAT[768] c0bbd5877662
val_0 INT64[1] 12a3ae445661
val_1 FLOAT[] 6708d9be4956
val_10 FLOAT[768,2304] db5df112c9e6
val_11 FLOAT[768,3072] 4ce00dc9f7a8
val_12 FLOAT[3072,768] 307e87ad9223
val_13 FLOAT[768,2304] bf9292631d32
val_14 FLOAT[768,3072] 10362b7c4747
val_15 FLOAT[3072,768] 1651c30aa716
val_16 FLOAT[768,2304] d5405ebd0e99
val_17 FLOAT[768,3072] 2f18af5c9632
val_18 FLOAT[3072,768] 1ecbd70008c0
val_19 FLOAT[768,2304] 45cf3ecf6c7c
val_2 INT64[1] 7c9fa136d441
val_20 FLOAT[768,3072] f8839c9153d1
val_21 FLOAT[3072,768] 0176acc99d8b
val_22 FLOAT[768,2304] 705ad19f914d
val_23 FLOAT[768,3072] c8a0a4a40aec
val_24 FLOAT[3072,768] 98aaad431455
val_25 FLOAT[768,2304] 6bb06480a085
val_26 FLOAT[768,3072] 0b8ba90d2f42
val_27 FLOAT[3072,768] be68aee42e1a
val_28 FLOAT[768,2304] 82d236054403
val_29 FLOAT[768,3072] 722d941b90bd
val_3 INT64[1] 6a69a6cc7473
val_30 FLOAT[3072,768] 1c360472662c
val_31 FLOAT[768,2304] b0e73cb95be5
val_32 FLOAT[768,3072] 3d67402e1d4f
val_33 FLOAT[3072,768] c17d2e4ae58f
val_34 FLOAT[768,2304] 5e1c8f0649ce
val_35 FLOAT[768,3072] 542f70f9ac8b
val_36 FLOAT[3072,768] 127fb672ebb0
val_37 FLOAT[768,2304] 124f886c1d94
val_38 FLOAT[768,3072] 5b3c170d2b8d
val_39 FLOAT[3072,768] e8c3488791a6
val_4 FLOAT[768,2304] 6270ccea47d1
val_5 FLOAT[768,3072] 251677b853aa
val_6 FLOAT[3072,768] ac2711e86fe1
val_7 FLOAT[768,2304] 5c3cca16470e
val_8 FLOAT[768,3072] c8450fdb9f3c
val_9 FLOAT[3072,768] 6387d9273663
