Upload LSTLlamaForCausalLM

#1
by tennant - opened
config.json ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": ".",
3
+ "add_attn": true,
4
+ "architectures": [
5
+ "LSTLlamaForCausalLM"
6
+ ],
7
+ "attention_bias": false,
8
+ "attn_reduction": 4,
9
+ "bos_token_id": 1,
10
+ "eos_token_id": 2,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 5120,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 13824,
15
+ "max_position_embeddings": 4096,
16
+ "mlp_r": 16,
17
+ "model_type": "lst",
18
+ "num_attention_heads": 40,
19
+ "num_hidden_layers": 40,
20
+ "num_key_value_heads": 40,
21
+ "pretraining_tp": 1,
22
+ "rms_norm_eps": 1e-05,
23
+ "rope_scaling": null,
24
+ "rope_theta": 10000.0,
25
+ "tie_word_embeddings": false,
26
+ "torch_dtype": "float32",
27
+ "transformers_version": "4.34.0",
28
+ "use_cache": false,
29
+ "vocab_size": 32000
30
+ }
generation_config.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 1,
3
+ "do_sample": true,
4
+ "eos_token_id": 2,
5
+ "max_length": 4096,
6
+ "pad_token_id": 0,
7
+ "temperature": 0.6,
8
+ "top_p": 0.9,
9
+ "transformers_version": "4.34.0"
10
+ }
pytorch_model-00001-of-00006.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3b394e404757584e15e8bf8e30d8cff384e2f77146512a9f7f745e02684cd2a5
3
+ size 9956539570
pytorch_model-00002-of-00006.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c3625ee960bf2cc6985a789b746c273c3e9da51cdebd0648fdcb1611fc22c3d5
3
+ size 9940852630
pytorch_model-00003-of-00006.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2c570c83f90738c621803a90b58bff5a3aa3f896b55896640638c4ada55640f9
3
+ size 9940852642
pytorch_model-00004-of-00006.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6caaf739341c6c5517e69a478c9c8132a197ca4e460f0d1abfc3c2318625f26c
3
+ size 9867410976
pytorch_model-00005-of-00006.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:89f0029b24376040e220d12f199f730d09946215a9e3c4e6f24b5e581bb5a996
3
+ size 9867452612
pytorch_model-00006-of-00006.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:269d70aff91dd41f102887354b625227a1967761bb331f90edc22169aeaddad1
3
+ size 5665241719
pytorch_model.bin.index.json ADDED
@@ -0,0 +1,850 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "metadata": {
3
+ "total_size": 55238062080
4
+ },
5
+ "weight_map": {
6
+ "lm_head.weight": "pytorch_model-00006-of-00006.bin",
7
+ "model.added_attn.0.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
8
+ "model.added_attn.0.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
9
+ "model.added_attn.0.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
10
+ "model.added_attn.0.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
11
+ "model.added_attn.0.input_proj.bias": "pytorch_model-00006-of-00006.bin",
12
+ "model.added_attn.0.input_proj.weight": "pytorch_model-00006-of-00006.bin",
13
+ "model.added_attn.0.output_proj.bias": "pytorch_model-00006-of-00006.bin",
14
+ "model.added_attn.0.output_proj.weight": "pytorch_model-00006-of-00006.bin",
15
+ "model.added_attn.1.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
16
+ "model.added_attn.1.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
17
+ "model.added_attn.1.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
18
+ "model.added_attn.1.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
19
+ "model.added_attn.1.input_proj.bias": "pytorch_model-00006-of-00006.bin",
20
+ "model.added_attn.1.input_proj.weight": "pytorch_model-00006-of-00006.bin",
21
+ "model.added_attn.1.output_proj.bias": "pytorch_model-00006-of-00006.bin",
22
+ "model.added_attn.1.output_proj.weight": "pytorch_model-00006-of-00006.bin",
23
+ "model.added_attn.10.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
24
+ "model.added_attn.10.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
25
+ "model.added_attn.10.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
26
+ "model.added_attn.10.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
27
+ "model.added_attn.10.input_proj.bias": "pytorch_model-00006-of-00006.bin",
28
+ "model.added_attn.10.input_proj.weight": "pytorch_model-00006-of-00006.bin",
29
+ "model.added_attn.10.output_proj.bias": "pytorch_model-00006-of-00006.bin",
30
+ "model.added_attn.10.output_proj.weight": "pytorch_model-00006-of-00006.bin",
31
+ "model.added_attn.11.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
32
+ "model.added_attn.11.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
33
+ "model.added_attn.11.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
34
+ "model.added_attn.11.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
35
+ "model.added_attn.11.input_proj.bias": "pytorch_model-00006-of-00006.bin",
36
+ "model.added_attn.11.input_proj.weight": "pytorch_model-00006-of-00006.bin",
37
+ "model.added_attn.11.output_proj.bias": "pytorch_model-00006-of-00006.bin",
38
+ "model.added_attn.11.output_proj.weight": "pytorch_model-00006-of-00006.bin",
39
+ "model.added_attn.12.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
40
+ "model.added_attn.12.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
41
+ "model.added_attn.12.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
42
+ "model.added_attn.12.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
43
+ "model.added_attn.12.input_proj.bias": "pytorch_model-00006-of-00006.bin",
44
+ "model.added_attn.12.input_proj.weight": "pytorch_model-00006-of-00006.bin",
45
+ "model.added_attn.12.output_proj.bias": "pytorch_model-00006-of-00006.bin",
46
+ "model.added_attn.12.output_proj.weight": "pytorch_model-00006-of-00006.bin",
47
+ "model.added_attn.13.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
48
+ "model.added_attn.13.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
49
+ "model.added_attn.13.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
50
+ "model.added_attn.13.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
51
+ "model.added_attn.13.input_proj.bias": "pytorch_model-00006-of-00006.bin",
52
+ "model.added_attn.13.input_proj.weight": "pytorch_model-00006-of-00006.bin",
53
+ "model.added_attn.13.output_proj.bias": "pytorch_model-00006-of-00006.bin",
54
+ "model.added_attn.13.output_proj.weight": "pytorch_model-00006-of-00006.bin",
55
+ "model.added_attn.14.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
56
+ "model.added_attn.14.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
57
+ "model.added_attn.14.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
58
+ "model.added_attn.14.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
59
+ "model.added_attn.14.input_proj.bias": "pytorch_model-00006-of-00006.bin",
60
+ "model.added_attn.14.input_proj.weight": "pytorch_model-00006-of-00006.bin",
61
+ "model.added_attn.14.output_proj.bias": "pytorch_model-00006-of-00006.bin",
62
+ "model.added_attn.14.output_proj.weight": "pytorch_model-00006-of-00006.bin",
63
+ "model.added_attn.15.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
64
+ "model.added_attn.15.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
65
+ "model.added_attn.15.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
66
+ "model.added_attn.15.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
67
+ "model.added_attn.15.input_proj.bias": "pytorch_model-00006-of-00006.bin",
68
+ "model.added_attn.15.input_proj.weight": "pytorch_model-00006-of-00006.bin",
69
+ "model.added_attn.15.output_proj.bias": "pytorch_model-00006-of-00006.bin",
70
+ "model.added_attn.15.output_proj.weight": "pytorch_model-00006-of-00006.bin",
71
+ "model.added_attn.16.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
72
+ "model.added_attn.16.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
73
+ "model.added_attn.16.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
74
+ "model.added_attn.16.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
75
+ "model.added_attn.16.input_proj.bias": "pytorch_model-00006-of-00006.bin",
76
+ "model.added_attn.16.input_proj.weight": "pytorch_model-00006-of-00006.bin",
77
+ "model.added_attn.16.output_proj.bias": "pytorch_model-00006-of-00006.bin",
78
+ "model.added_attn.16.output_proj.weight": "pytorch_model-00006-of-00006.bin",
79
+ "model.added_attn.17.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
80
+ "model.added_attn.17.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
81
+ "model.added_attn.17.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
82
+ "model.added_attn.17.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
83
+ "model.added_attn.17.input_proj.bias": "pytorch_model-00006-of-00006.bin",
84
+ "model.added_attn.17.input_proj.weight": "pytorch_model-00006-of-00006.bin",
85
+ "model.added_attn.17.output_proj.bias": "pytorch_model-00006-of-00006.bin",
86
+ "model.added_attn.17.output_proj.weight": "pytorch_model-00006-of-00006.bin",
87
+ "model.added_attn.18.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
88
+ "model.added_attn.18.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
89
+ "model.added_attn.18.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
90
+ "model.added_attn.18.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
91
+ "model.added_attn.18.input_proj.bias": "pytorch_model-00006-of-00006.bin",
92
+ "model.added_attn.18.input_proj.weight": "pytorch_model-00006-of-00006.bin",
93
+ "model.added_attn.18.output_proj.bias": "pytorch_model-00006-of-00006.bin",
94
+ "model.added_attn.18.output_proj.weight": "pytorch_model-00006-of-00006.bin",
95
+ "model.added_attn.19.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
96
+ "model.added_attn.19.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
97
+ "model.added_attn.19.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
98
+ "model.added_attn.19.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
99
+ "model.added_attn.19.input_proj.bias": "pytorch_model-00006-of-00006.bin",
100
+ "model.added_attn.19.input_proj.weight": "pytorch_model-00006-of-00006.bin",
101
+ "model.added_attn.19.output_proj.bias": "pytorch_model-00006-of-00006.bin",
102
+ "model.added_attn.19.output_proj.weight": "pytorch_model-00006-of-00006.bin",
103
+ "model.added_attn.2.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
104
+ "model.added_attn.2.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
105
+ "model.added_attn.2.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
106
+ "model.added_attn.2.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
107
+ "model.added_attn.2.input_proj.bias": "pytorch_model-00006-of-00006.bin",
108
+ "model.added_attn.2.input_proj.weight": "pytorch_model-00006-of-00006.bin",
109
+ "model.added_attn.2.output_proj.bias": "pytorch_model-00006-of-00006.bin",
110
+ "model.added_attn.2.output_proj.weight": "pytorch_model-00006-of-00006.bin",
111
+ "model.added_attn.20.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
112
+ "model.added_attn.20.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
113
+ "model.added_attn.20.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
114
+ "model.added_attn.20.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
115
+ "model.added_attn.20.input_proj.bias": "pytorch_model-00006-of-00006.bin",
116
+ "model.added_attn.20.input_proj.weight": "pytorch_model-00006-of-00006.bin",
117
+ "model.added_attn.20.output_proj.bias": "pytorch_model-00006-of-00006.bin",
118
+ "model.added_attn.20.output_proj.weight": "pytorch_model-00006-of-00006.bin",
119
+ "model.added_attn.21.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
120
+ "model.added_attn.21.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
121
+ "model.added_attn.21.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
122
+ "model.added_attn.21.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
123
+ "model.added_attn.21.input_proj.bias": "pytorch_model-00006-of-00006.bin",
124
+ "model.added_attn.21.input_proj.weight": "pytorch_model-00006-of-00006.bin",
125
+ "model.added_attn.21.output_proj.bias": "pytorch_model-00006-of-00006.bin",
126
+ "model.added_attn.21.output_proj.weight": "pytorch_model-00006-of-00006.bin",
127
+ "model.added_attn.22.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
128
+ "model.added_attn.22.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
129
+ "model.added_attn.22.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
130
+ "model.added_attn.22.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
131
+ "model.added_attn.22.input_proj.bias": "pytorch_model-00006-of-00006.bin",
132
+ "model.added_attn.22.input_proj.weight": "pytorch_model-00006-of-00006.bin",
133
+ "model.added_attn.22.output_proj.bias": "pytorch_model-00006-of-00006.bin",
134
+ "model.added_attn.22.output_proj.weight": "pytorch_model-00006-of-00006.bin",
135
+ "model.added_attn.23.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
136
+ "model.added_attn.23.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
137
+ "model.added_attn.23.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
138
+ "model.added_attn.23.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
139
+ "model.added_attn.23.input_proj.bias": "pytorch_model-00006-of-00006.bin",
140
+ "model.added_attn.23.input_proj.weight": "pytorch_model-00006-of-00006.bin",
141
+ "model.added_attn.23.output_proj.bias": "pytorch_model-00006-of-00006.bin",
142
+ "model.added_attn.23.output_proj.weight": "pytorch_model-00006-of-00006.bin",
143
+ "model.added_attn.24.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
144
+ "model.added_attn.24.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
145
+ "model.added_attn.24.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
146
+ "model.added_attn.24.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
147
+ "model.added_attn.24.input_proj.bias": "pytorch_model-00006-of-00006.bin",
148
+ "model.added_attn.24.input_proj.weight": "pytorch_model-00006-of-00006.bin",
149
+ "model.added_attn.24.output_proj.bias": "pytorch_model-00006-of-00006.bin",
150
+ "model.added_attn.24.output_proj.weight": "pytorch_model-00006-of-00006.bin",
151
+ "model.added_attn.25.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
152
+ "model.added_attn.25.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
153
+ "model.added_attn.25.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
154
+ "model.added_attn.25.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
155
+ "model.added_attn.25.input_proj.bias": "pytorch_model-00006-of-00006.bin",
156
+ "model.added_attn.25.input_proj.weight": "pytorch_model-00006-of-00006.bin",
157
+ "model.added_attn.25.output_proj.bias": "pytorch_model-00006-of-00006.bin",
158
+ "model.added_attn.25.output_proj.weight": "pytorch_model-00006-of-00006.bin",
159
+ "model.added_attn.26.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
160
+ "model.added_attn.26.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
161
+ "model.added_attn.26.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
162
+ "model.added_attn.26.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
163
+ "model.added_attn.26.input_proj.bias": "pytorch_model-00006-of-00006.bin",
164
+ "model.added_attn.26.input_proj.weight": "pytorch_model-00006-of-00006.bin",
165
+ "model.added_attn.26.output_proj.bias": "pytorch_model-00006-of-00006.bin",
166
+ "model.added_attn.26.output_proj.weight": "pytorch_model-00006-of-00006.bin",
167
+ "model.added_attn.27.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
168
+ "model.added_attn.27.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
169
+ "model.added_attn.27.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
170
+ "model.added_attn.27.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
171
+ "model.added_attn.27.input_proj.bias": "pytorch_model-00006-of-00006.bin",
172
+ "model.added_attn.27.input_proj.weight": "pytorch_model-00006-of-00006.bin",
173
+ "model.added_attn.27.output_proj.bias": "pytorch_model-00006-of-00006.bin",
174
+ "model.added_attn.27.output_proj.weight": "pytorch_model-00006-of-00006.bin",
175
+ "model.added_attn.28.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
176
+ "model.added_attn.28.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
177
+ "model.added_attn.28.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
178
+ "model.added_attn.28.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
179
+ "model.added_attn.28.input_proj.bias": "pytorch_model-00006-of-00006.bin",
180
+ "model.added_attn.28.input_proj.weight": "pytorch_model-00006-of-00006.bin",
181
+ "model.added_attn.28.output_proj.bias": "pytorch_model-00006-of-00006.bin",
182
+ "model.added_attn.28.output_proj.weight": "pytorch_model-00006-of-00006.bin",
183
+ "model.added_attn.29.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
184
+ "model.added_attn.29.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
185
+ "model.added_attn.29.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
186
+ "model.added_attn.29.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
187
+ "model.added_attn.29.input_proj.bias": "pytorch_model-00006-of-00006.bin",
188
+ "model.added_attn.29.input_proj.weight": "pytorch_model-00006-of-00006.bin",
189
+ "model.added_attn.29.output_proj.bias": "pytorch_model-00006-of-00006.bin",
190
+ "model.added_attn.29.output_proj.weight": "pytorch_model-00006-of-00006.bin",
191
+ "model.added_attn.3.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
192
+ "model.added_attn.3.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
193
+ "model.added_attn.3.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
194
+ "model.added_attn.3.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
195
+ "model.added_attn.3.input_proj.bias": "pytorch_model-00006-of-00006.bin",
196
+ "model.added_attn.3.input_proj.weight": "pytorch_model-00006-of-00006.bin",
197
+ "model.added_attn.3.output_proj.bias": "pytorch_model-00006-of-00006.bin",
198
+ "model.added_attn.3.output_proj.weight": "pytorch_model-00006-of-00006.bin",
199
+ "model.added_attn.30.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
200
+ "model.added_attn.30.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
201
+ "model.added_attn.30.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
202
+ "model.added_attn.30.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
203
+ "model.added_attn.30.input_proj.bias": "pytorch_model-00006-of-00006.bin",
204
+ "model.added_attn.30.input_proj.weight": "pytorch_model-00006-of-00006.bin",
205
+ "model.added_attn.30.output_proj.bias": "pytorch_model-00006-of-00006.bin",
206
+ "model.added_attn.30.output_proj.weight": "pytorch_model-00006-of-00006.bin",
207
+ "model.added_attn.31.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
208
+ "model.added_attn.31.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
209
+ "model.added_attn.31.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
210
+ "model.added_attn.31.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
211
+ "model.added_attn.31.input_proj.bias": "pytorch_model-00006-of-00006.bin",
212
+ "model.added_attn.31.input_proj.weight": "pytorch_model-00006-of-00006.bin",
213
+ "model.added_attn.31.output_proj.bias": "pytorch_model-00006-of-00006.bin",
214
+ "model.added_attn.31.output_proj.weight": "pytorch_model-00006-of-00006.bin",
215
+ "model.added_attn.32.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
216
+ "model.added_attn.32.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
217
+ "model.added_attn.32.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
218
+ "model.added_attn.32.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
219
+ "model.added_attn.32.input_proj.bias": "pytorch_model-00006-of-00006.bin",
220
+ "model.added_attn.32.input_proj.weight": "pytorch_model-00006-of-00006.bin",
221
+ "model.added_attn.32.output_proj.bias": "pytorch_model-00006-of-00006.bin",
222
+ "model.added_attn.32.output_proj.weight": "pytorch_model-00006-of-00006.bin",
223
+ "model.added_attn.33.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
224
+ "model.added_attn.33.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
225
+ "model.added_attn.33.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
226
+ "model.added_attn.33.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
227
+ "model.added_attn.33.input_proj.bias": "pytorch_model-00006-of-00006.bin",
228
+ "model.added_attn.33.input_proj.weight": "pytorch_model-00006-of-00006.bin",
229
+ "model.added_attn.33.output_proj.bias": "pytorch_model-00006-of-00006.bin",
230
+ "model.added_attn.33.output_proj.weight": "pytorch_model-00006-of-00006.bin",
231
+ "model.added_attn.34.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
232
+ "model.added_attn.34.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
233
+ "model.added_attn.34.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
234
+ "model.added_attn.34.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
235
+ "model.added_attn.34.input_proj.bias": "pytorch_model-00006-of-00006.bin",
236
+ "model.added_attn.34.input_proj.weight": "pytorch_model-00006-of-00006.bin",
237
+ "model.added_attn.34.output_proj.bias": "pytorch_model-00006-of-00006.bin",
238
+ "model.added_attn.34.output_proj.weight": "pytorch_model-00006-of-00006.bin",
239
+ "model.added_attn.35.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
240
+ "model.added_attn.35.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
241
+ "model.added_attn.35.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
242
+ "model.added_attn.35.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
243
+ "model.added_attn.35.input_proj.bias": "pytorch_model-00006-of-00006.bin",
244
+ "model.added_attn.35.input_proj.weight": "pytorch_model-00006-of-00006.bin",
245
+ "model.added_attn.35.output_proj.bias": "pytorch_model-00006-of-00006.bin",
246
+ "model.added_attn.35.output_proj.weight": "pytorch_model-00006-of-00006.bin",
247
+ "model.added_attn.36.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
248
+ "model.added_attn.36.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
249
+ "model.added_attn.36.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
250
+ "model.added_attn.36.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
251
+ "model.added_attn.36.input_proj.bias": "pytorch_model-00006-of-00006.bin",
252
+ "model.added_attn.36.input_proj.weight": "pytorch_model-00006-of-00006.bin",
253
+ "model.added_attn.36.output_proj.bias": "pytorch_model-00006-of-00006.bin",
254
+ "model.added_attn.36.output_proj.weight": "pytorch_model-00006-of-00006.bin",
255
+ "model.added_attn.37.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
256
+ "model.added_attn.37.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
257
+ "model.added_attn.37.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
258
+ "model.added_attn.37.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
259
+ "model.added_attn.37.input_proj.bias": "pytorch_model-00006-of-00006.bin",
260
+ "model.added_attn.37.input_proj.weight": "pytorch_model-00006-of-00006.bin",
261
+ "model.added_attn.37.output_proj.bias": "pytorch_model-00006-of-00006.bin",
262
+ "model.added_attn.37.output_proj.weight": "pytorch_model-00006-of-00006.bin",
263
+ "model.added_attn.38.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
264
+ "model.added_attn.38.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
265
+ "model.added_attn.38.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
266
+ "model.added_attn.38.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
267
+ "model.added_attn.38.input_proj.bias": "pytorch_model-00006-of-00006.bin",
268
+ "model.added_attn.38.input_proj.weight": "pytorch_model-00006-of-00006.bin",
269
+ "model.added_attn.38.output_proj.bias": "pytorch_model-00006-of-00006.bin",
270
+ "model.added_attn.38.output_proj.weight": "pytorch_model-00006-of-00006.bin",
271
+ "model.added_attn.39.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
272
+ "model.added_attn.39.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
273
+ "model.added_attn.39.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
274
+ "model.added_attn.39.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
275
+ "model.added_attn.39.input_proj.bias": "pytorch_model-00006-of-00006.bin",
276
+ "model.added_attn.39.input_proj.weight": "pytorch_model-00006-of-00006.bin",
277
+ "model.added_attn.39.output_proj.bias": "pytorch_model-00006-of-00006.bin",
278
+ "model.added_attn.39.output_proj.weight": "pytorch_model-00006-of-00006.bin",
279
+ "model.added_attn.4.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
280
+ "model.added_attn.4.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
281
+ "model.added_attn.4.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
282
+ "model.added_attn.4.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
283
+ "model.added_attn.4.input_proj.bias": "pytorch_model-00006-of-00006.bin",
284
+ "model.added_attn.4.input_proj.weight": "pytorch_model-00006-of-00006.bin",
285
+ "model.added_attn.4.output_proj.bias": "pytorch_model-00006-of-00006.bin",
286
+ "model.added_attn.4.output_proj.weight": "pytorch_model-00006-of-00006.bin",
287
+ "model.added_attn.5.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
288
+ "model.added_attn.5.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
289
+ "model.added_attn.5.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
290
+ "model.added_attn.5.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
291
+ "model.added_attn.5.input_proj.bias": "pytorch_model-00006-of-00006.bin",
292
+ "model.added_attn.5.input_proj.weight": "pytorch_model-00006-of-00006.bin",
293
+ "model.added_attn.5.output_proj.bias": "pytorch_model-00006-of-00006.bin",
294
+ "model.added_attn.5.output_proj.weight": "pytorch_model-00006-of-00006.bin",
295
+ "model.added_attn.6.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
296
+ "model.added_attn.6.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
297
+ "model.added_attn.6.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
298
+ "model.added_attn.6.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
299
+ "model.added_attn.6.input_proj.bias": "pytorch_model-00006-of-00006.bin",
300
+ "model.added_attn.6.input_proj.weight": "pytorch_model-00006-of-00006.bin",
301
+ "model.added_attn.6.output_proj.bias": "pytorch_model-00006-of-00006.bin",
302
+ "model.added_attn.6.output_proj.weight": "pytorch_model-00006-of-00006.bin",
303
+ "model.added_attn.7.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
304
+ "model.added_attn.7.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
305
+ "model.added_attn.7.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
306
+ "model.added_attn.7.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
307
+ "model.added_attn.7.input_proj.bias": "pytorch_model-00006-of-00006.bin",
308
+ "model.added_attn.7.input_proj.weight": "pytorch_model-00006-of-00006.bin",
309
+ "model.added_attn.7.output_proj.bias": "pytorch_model-00006-of-00006.bin",
310
+ "model.added_attn.7.output_proj.weight": "pytorch_model-00006-of-00006.bin",
311
+ "model.added_attn.8.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
312
+ "model.added_attn.8.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
313
+ "model.added_attn.8.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
314
+ "model.added_attn.8.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
315
+ "model.added_attn.8.input_proj.bias": "pytorch_model-00006-of-00006.bin",
316
+ "model.added_attn.8.input_proj.weight": "pytorch_model-00006-of-00006.bin",
317
+ "model.added_attn.8.output_proj.bias": "pytorch_model-00006-of-00006.bin",
318
+ "model.added_attn.8.output_proj.weight": "pytorch_model-00006-of-00006.bin",
319
+ "model.added_attn.9.attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
320
+ "model.added_attn.9.attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
321
+ "model.added_attn.9.attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
322
+ "model.added_attn.9.attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
323
+ "model.added_attn.9.input_proj.bias": "pytorch_model-00006-of-00006.bin",
324
+ "model.added_attn.9.input_proj.weight": "pytorch_model-00006-of-00006.bin",
325
+ "model.added_attn.9.output_proj.bias": "pytorch_model-00006-of-00006.bin",
326
+ "model.added_attn.9.output_proj.weight": "pytorch_model-00006-of-00006.bin",
327
+ "model.added_lns.0.weight": "pytorch_model-00006-of-00006.bin",
328
+ "model.added_lns.1.weight": "pytorch_model-00006-of-00006.bin",
329
+ "model.added_lns.10.weight": "pytorch_model-00006-of-00006.bin",
330
+ "model.added_lns.11.weight": "pytorch_model-00006-of-00006.bin",
331
+ "model.added_lns.12.weight": "pytorch_model-00006-of-00006.bin",
332
+ "model.added_lns.13.weight": "pytorch_model-00006-of-00006.bin",
333
+ "model.added_lns.14.weight": "pytorch_model-00006-of-00006.bin",
334
+ "model.added_lns.15.weight": "pytorch_model-00006-of-00006.bin",
335
+ "model.added_lns.16.weight": "pytorch_model-00006-of-00006.bin",
336
+ "model.added_lns.17.weight": "pytorch_model-00006-of-00006.bin",
337
+ "model.added_lns.18.weight": "pytorch_model-00006-of-00006.bin",
338
+ "model.added_lns.19.weight": "pytorch_model-00006-of-00006.bin",
339
+ "model.added_lns.2.weight": "pytorch_model-00006-of-00006.bin",
340
+ "model.added_lns.20.weight": "pytorch_model-00006-of-00006.bin",
341
+ "model.added_lns.21.weight": "pytorch_model-00006-of-00006.bin",
342
+ "model.added_lns.22.weight": "pytorch_model-00006-of-00006.bin",
343
+ "model.added_lns.23.weight": "pytorch_model-00006-of-00006.bin",
344
+ "model.added_lns.24.weight": "pytorch_model-00006-of-00006.bin",
345
+ "model.added_lns.25.weight": "pytorch_model-00006-of-00006.bin",
346
+ "model.added_lns.26.weight": "pytorch_model-00006-of-00006.bin",
347
+ "model.added_lns.27.weight": "pytorch_model-00006-of-00006.bin",
348
+ "model.added_lns.28.weight": "pytorch_model-00006-of-00006.bin",
349
+ "model.added_lns.29.weight": "pytorch_model-00006-of-00006.bin",
350
+ "model.added_lns.3.weight": "pytorch_model-00006-of-00006.bin",
351
+ "model.added_lns.30.weight": "pytorch_model-00006-of-00006.bin",
352
+ "model.added_lns.31.weight": "pytorch_model-00006-of-00006.bin",
353
+ "model.added_lns.32.weight": "pytorch_model-00006-of-00006.bin",
354
+ "model.added_lns.33.weight": "pytorch_model-00006-of-00006.bin",
355
+ "model.added_lns.34.weight": "pytorch_model-00006-of-00006.bin",
356
+ "model.added_lns.35.weight": "pytorch_model-00006-of-00006.bin",
357
+ "model.added_lns.36.weight": "pytorch_model-00006-of-00006.bin",
358
+ "model.added_lns.37.weight": "pytorch_model-00006-of-00006.bin",
359
+ "model.added_lns.38.weight": "pytorch_model-00006-of-00006.bin",
360
+ "model.added_lns.39.weight": "pytorch_model-00006-of-00006.bin",
361
+ "model.added_lns.4.weight": "pytorch_model-00006-of-00006.bin",
362
+ "model.added_lns.5.weight": "pytorch_model-00006-of-00006.bin",
363
+ "model.added_lns.6.weight": "pytorch_model-00006-of-00006.bin",
364
+ "model.added_lns.7.weight": "pytorch_model-00006-of-00006.bin",
365
+ "model.added_lns.8.weight": "pytorch_model-00006-of-00006.bin",
366
+ "model.added_lns.9.weight": "pytorch_model-00006-of-00006.bin",
367
+ "model.added_lrmlp.0.mm_a.weight": "pytorch_model-00006-of-00006.bin",
368
+ "model.added_lrmlp.0.mm_b.weight": "pytorch_model-00006-of-00006.bin",
369
+ "model.added_lrmlp.1.mm_a.weight": "pytorch_model-00006-of-00006.bin",
370
+ "model.added_lrmlp.1.mm_b.weight": "pytorch_model-00006-of-00006.bin",
371
+ "model.added_lrmlp.10.mm_a.weight": "pytorch_model-00006-of-00006.bin",
372
+ "model.added_lrmlp.10.mm_b.weight": "pytorch_model-00006-of-00006.bin",
373
+ "model.added_lrmlp.11.mm_a.weight": "pytorch_model-00006-of-00006.bin",
374
+ "model.added_lrmlp.11.mm_b.weight": "pytorch_model-00006-of-00006.bin",
375
+ "model.added_lrmlp.12.mm_a.weight": "pytorch_model-00006-of-00006.bin",
376
+ "model.added_lrmlp.12.mm_b.weight": "pytorch_model-00006-of-00006.bin",
377
+ "model.added_lrmlp.13.mm_a.weight": "pytorch_model-00006-of-00006.bin",
378
+ "model.added_lrmlp.13.mm_b.weight": "pytorch_model-00006-of-00006.bin",
379
+ "model.added_lrmlp.14.mm_a.weight": "pytorch_model-00006-of-00006.bin",
380
+ "model.added_lrmlp.14.mm_b.weight": "pytorch_model-00006-of-00006.bin",
381
+ "model.added_lrmlp.15.mm_a.weight": "pytorch_model-00006-of-00006.bin",
382
+ "model.added_lrmlp.15.mm_b.weight": "pytorch_model-00006-of-00006.bin",
383
+ "model.added_lrmlp.16.mm_a.weight": "pytorch_model-00006-of-00006.bin",
384
+ "model.added_lrmlp.16.mm_b.weight": "pytorch_model-00006-of-00006.bin",
385
+ "model.added_lrmlp.17.mm_a.weight": "pytorch_model-00006-of-00006.bin",
386
+ "model.added_lrmlp.17.mm_b.weight": "pytorch_model-00006-of-00006.bin",
387
+ "model.added_lrmlp.18.mm_a.weight": "pytorch_model-00006-of-00006.bin",
388
+ "model.added_lrmlp.18.mm_b.weight": "pytorch_model-00006-of-00006.bin",
389
+ "model.added_lrmlp.19.mm_a.weight": "pytorch_model-00006-of-00006.bin",
390
+ "model.added_lrmlp.19.mm_b.weight": "pytorch_model-00006-of-00006.bin",
391
+ "model.added_lrmlp.2.mm_a.weight": "pytorch_model-00006-of-00006.bin",
392
+ "model.added_lrmlp.2.mm_b.weight": "pytorch_model-00006-of-00006.bin",
393
+ "model.added_lrmlp.20.mm_a.weight": "pytorch_model-00006-of-00006.bin",
394
+ "model.added_lrmlp.20.mm_b.weight": "pytorch_model-00006-of-00006.bin",
395
+ "model.added_lrmlp.21.mm_a.weight": "pytorch_model-00006-of-00006.bin",
396
+ "model.added_lrmlp.21.mm_b.weight": "pytorch_model-00006-of-00006.bin",
397
+ "model.added_lrmlp.22.mm_a.weight": "pytorch_model-00006-of-00006.bin",
398
+ "model.added_lrmlp.22.mm_b.weight": "pytorch_model-00006-of-00006.bin",
399
+ "model.added_lrmlp.23.mm_a.weight": "pytorch_model-00006-of-00006.bin",
400
+ "model.added_lrmlp.23.mm_b.weight": "pytorch_model-00006-of-00006.bin",
401
+ "model.added_lrmlp.24.mm_a.weight": "pytorch_model-00006-of-00006.bin",
402
+ "model.added_lrmlp.24.mm_b.weight": "pytorch_model-00006-of-00006.bin",
403
+ "model.added_lrmlp.25.mm_a.weight": "pytorch_model-00006-of-00006.bin",
404
+ "model.added_lrmlp.25.mm_b.weight": "pytorch_model-00006-of-00006.bin",
405
+ "model.added_lrmlp.26.mm_a.weight": "pytorch_model-00006-of-00006.bin",
406
+ "model.added_lrmlp.26.mm_b.weight": "pytorch_model-00006-of-00006.bin",
407
+ "model.added_lrmlp.27.mm_a.weight": "pytorch_model-00006-of-00006.bin",
408
+ "model.added_lrmlp.27.mm_b.weight": "pytorch_model-00006-of-00006.bin",
409
+ "model.added_lrmlp.28.mm_a.weight": "pytorch_model-00006-of-00006.bin",
410
+ "model.added_lrmlp.28.mm_b.weight": "pytorch_model-00006-of-00006.bin",
411
+ "model.added_lrmlp.29.mm_a.weight": "pytorch_model-00006-of-00006.bin",
412
+ "model.added_lrmlp.29.mm_b.weight": "pytorch_model-00006-of-00006.bin",
413
+ "model.added_lrmlp.3.mm_a.weight": "pytorch_model-00006-of-00006.bin",
414
+ "model.added_lrmlp.3.mm_b.weight": "pytorch_model-00006-of-00006.bin",
415
+ "model.added_lrmlp.30.mm_a.weight": "pytorch_model-00006-of-00006.bin",
416
+ "model.added_lrmlp.30.mm_b.weight": "pytorch_model-00006-of-00006.bin",
417
+ "model.added_lrmlp.31.mm_a.weight": "pytorch_model-00006-of-00006.bin",
418
+ "model.added_lrmlp.31.mm_b.weight": "pytorch_model-00006-of-00006.bin",
419
+ "model.added_lrmlp.32.mm_a.weight": "pytorch_model-00006-of-00006.bin",
420
+ "model.added_lrmlp.32.mm_b.weight": "pytorch_model-00006-of-00006.bin",
421
+ "model.added_lrmlp.33.mm_a.weight": "pytorch_model-00006-of-00006.bin",
422
+ "model.added_lrmlp.33.mm_b.weight": "pytorch_model-00006-of-00006.bin",
423
+ "model.added_lrmlp.34.mm_a.weight": "pytorch_model-00006-of-00006.bin",
424
+ "model.added_lrmlp.34.mm_b.weight": "pytorch_model-00006-of-00006.bin",
425
+ "model.added_lrmlp.35.mm_a.weight": "pytorch_model-00006-of-00006.bin",
426
+ "model.added_lrmlp.35.mm_b.weight": "pytorch_model-00006-of-00006.bin",
427
+ "model.added_lrmlp.36.mm_a.weight": "pytorch_model-00006-of-00006.bin",
428
+ "model.added_lrmlp.36.mm_b.weight": "pytorch_model-00006-of-00006.bin",
429
+ "model.added_lrmlp.37.mm_a.weight": "pytorch_model-00006-of-00006.bin",
430
+ "model.added_lrmlp.37.mm_b.weight": "pytorch_model-00006-of-00006.bin",
431
+ "model.added_lrmlp.38.mm_a.weight": "pytorch_model-00006-of-00006.bin",
432
+ "model.added_lrmlp.38.mm_b.weight": "pytorch_model-00006-of-00006.bin",
433
+ "model.added_lrmlp.39.mm_a.weight": "pytorch_model-00006-of-00006.bin",
434
+ "model.added_lrmlp.39.mm_b.weight": "pytorch_model-00006-of-00006.bin",
435
+ "model.added_lrmlp.4.mm_a.weight": "pytorch_model-00006-of-00006.bin",
436
+ "model.added_lrmlp.4.mm_b.weight": "pytorch_model-00006-of-00006.bin",
437
+ "model.added_lrmlp.5.mm_a.weight": "pytorch_model-00006-of-00006.bin",
438
+ "model.added_lrmlp.5.mm_b.weight": "pytorch_model-00006-of-00006.bin",
439
+ "model.added_lrmlp.6.mm_a.weight": "pytorch_model-00006-of-00006.bin",
440
+ "model.added_lrmlp.6.mm_b.weight": "pytorch_model-00006-of-00006.bin",
441
+ "model.added_lrmlp.7.mm_a.weight": "pytorch_model-00006-of-00006.bin",
442
+ "model.added_lrmlp.7.mm_b.weight": "pytorch_model-00006-of-00006.bin",
443
+ "model.added_lrmlp.8.mm_a.weight": "pytorch_model-00006-of-00006.bin",
444
+ "model.added_lrmlp.8.mm_b.weight": "pytorch_model-00006-of-00006.bin",
445
+ "model.added_lrmlp.9.mm_a.weight": "pytorch_model-00006-of-00006.bin",
446
+ "model.added_lrmlp.9.mm_b.weight": "pytorch_model-00006-of-00006.bin",
447
+ "model.added_output_lns.0.weight": "pytorch_model-00006-of-00006.bin",
448
+ "model.added_output_lns.1.weight": "pytorch_model-00006-of-00006.bin",
449
+ "model.added_output_lns.10.weight": "pytorch_model-00006-of-00006.bin",
450
+ "model.added_output_lns.11.weight": "pytorch_model-00006-of-00006.bin",
451
+ "model.added_output_lns.12.weight": "pytorch_model-00006-of-00006.bin",
452
+ "model.added_output_lns.13.weight": "pytorch_model-00006-of-00006.bin",
453
+ "model.added_output_lns.14.weight": "pytorch_model-00006-of-00006.bin",
454
+ "model.added_output_lns.15.weight": "pytorch_model-00006-of-00006.bin",
455
+ "model.added_output_lns.16.weight": "pytorch_model-00006-of-00006.bin",
456
+ "model.added_output_lns.17.weight": "pytorch_model-00006-of-00006.bin",
457
+ "model.added_output_lns.18.weight": "pytorch_model-00006-of-00006.bin",
458
+ "model.added_output_lns.19.weight": "pytorch_model-00006-of-00006.bin",
459
+ "model.added_output_lns.2.weight": "pytorch_model-00006-of-00006.bin",
460
+ "model.added_output_lns.20.weight": "pytorch_model-00006-of-00006.bin",
461
+ "model.added_output_lns.21.weight": "pytorch_model-00006-of-00006.bin",
462
+ "model.added_output_lns.22.weight": "pytorch_model-00006-of-00006.bin",
463
+ "model.added_output_lns.23.weight": "pytorch_model-00006-of-00006.bin",
464
+ "model.added_output_lns.24.weight": "pytorch_model-00006-of-00006.bin",
465
+ "model.added_output_lns.25.weight": "pytorch_model-00006-of-00006.bin",
466
+ "model.added_output_lns.26.weight": "pytorch_model-00006-of-00006.bin",
467
+ "model.added_output_lns.27.weight": "pytorch_model-00006-of-00006.bin",
468
+ "model.added_output_lns.28.weight": "pytorch_model-00006-of-00006.bin",
469
+ "model.added_output_lns.29.weight": "pytorch_model-00006-of-00006.bin",
470
+ "model.added_output_lns.3.weight": "pytorch_model-00006-of-00006.bin",
471
+ "model.added_output_lns.30.weight": "pytorch_model-00006-of-00006.bin",
472
+ "model.added_output_lns.31.weight": "pytorch_model-00006-of-00006.bin",
473
+ "model.added_output_lns.32.weight": "pytorch_model-00006-of-00006.bin",
474
+ "model.added_output_lns.33.weight": "pytorch_model-00006-of-00006.bin",
475
+ "model.added_output_lns.34.weight": "pytorch_model-00006-of-00006.bin",
476
+ "model.added_output_lns.35.weight": "pytorch_model-00006-of-00006.bin",
477
+ "model.added_output_lns.36.weight": "pytorch_model-00006-of-00006.bin",
478
+ "model.added_output_lns.37.weight": "pytorch_model-00006-of-00006.bin",
479
+ "model.added_output_lns.38.weight": "pytorch_model-00006-of-00006.bin",
480
+ "model.added_output_lns.39.weight": "pytorch_model-00006-of-00006.bin",
481
+ "model.added_output_lns.4.weight": "pytorch_model-00006-of-00006.bin",
482
+ "model.added_output_lns.5.weight": "pytorch_model-00006-of-00006.bin",
483
+ "model.added_output_lns.6.weight": "pytorch_model-00006-of-00006.bin",
484
+ "model.added_output_lns.7.weight": "pytorch_model-00006-of-00006.bin",
485
+ "model.added_output_lns.8.weight": "pytorch_model-00006-of-00006.bin",
486
+ "model.added_output_lns.9.weight": "pytorch_model-00006-of-00006.bin",
487
+ "model.embed_tokens.weight": "pytorch_model-00001-of-00006.bin",
488
+ "model.layers.0.input_layernorm.weight": "pytorch_model-00001-of-00006.bin",
489
+ "model.layers.0.mlp.down_proj.weight": "pytorch_model-00001-of-00006.bin",
490
+ "model.layers.0.mlp.gate_proj.weight": "pytorch_model-00001-of-00006.bin",
491
+ "model.layers.0.mlp.up_proj.weight": "pytorch_model-00001-of-00006.bin",
492
+ "model.layers.0.post_attention_layernorm.weight": "pytorch_model-00001-of-00006.bin",
493
+ "model.layers.0.self_attn.k_proj.weight": "pytorch_model-00001-of-00006.bin",
494
+ "model.layers.0.self_attn.o_proj.weight": "pytorch_model-00001-of-00006.bin",
495
+ "model.layers.0.self_attn.q_proj.weight": "pytorch_model-00001-of-00006.bin",
496
+ "model.layers.0.self_attn.v_proj.weight": "pytorch_model-00001-of-00006.bin",
497
+ "model.layers.1.input_layernorm.weight": "pytorch_model-00001-of-00006.bin",
498
+ "model.layers.1.mlp.down_proj.weight": "pytorch_model-00001-of-00006.bin",
499
+ "model.layers.1.mlp.gate_proj.weight": "pytorch_model-00001-of-00006.bin",
500
+ "model.layers.1.mlp.up_proj.weight": "pytorch_model-00001-of-00006.bin",
501
+ "model.layers.1.post_attention_layernorm.weight": "pytorch_model-00001-of-00006.bin",
502
+ "model.layers.1.self_attn.k_proj.weight": "pytorch_model-00001-of-00006.bin",
503
+ "model.layers.1.self_attn.o_proj.weight": "pytorch_model-00001-of-00006.bin",
504
+ "model.layers.1.self_attn.q_proj.weight": "pytorch_model-00001-of-00006.bin",
505
+ "model.layers.1.self_attn.v_proj.weight": "pytorch_model-00001-of-00006.bin",
506
+ "model.layers.10.input_layernorm.weight": "pytorch_model-00002-of-00006.bin",
507
+ "model.layers.10.mlp.down_proj.weight": "pytorch_model-00002-of-00006.bin",
508
+ "model.layers.10.mlp.gate_proj.weight": "pytorch_model-00002-of-00006.bin",
509
+ "model.layers.10.mlp.up_proj.weight": "pytorch_model-00002-of-00006.bin",
510
+ "model.layers.10.post_attention_layernorm.weight": "pytorch_model-00002-of-00006.bin",
511
+ "model.layers.10.self_attn.k_proj.weight": "pytorch_model-00002-of-00006.bin",
512
+ "model.layers.10.self_attn.o_proj.weight": "pytorch_model-00002-of-00006.bin",
513
+ "model.layers.10.self_attn.q_proj.weight": "pytorch_model-00002-of-00006.bin",
514
+ "model.layers.10.self_attn.v_proj.weight": "pytorch_model-00002-of-00006.bin",
515
+ "model.layers.11.input_layernorm.weight": "pytorch_model-00002-of-00006.bin",
516
+ "model.layers.11.mlp.down_proj.weight": "pytorch_model-00002-of-00006.bin",
517
+ "model.layers.11.mlp.gate_proj.weight": "pytorch_model-00002-of-00006.bin",
518
+ "model.layers.11.mlp.up_proj.weight": "pytorch_model-00002-of-00006.bin",
519
+ "model.layers.11.post_attention_layernorm.weight": "pytorch_model-00002-of-00006.bin",
520
+ "model.layers.11.self_attn.k_proj.weight": "pytorch_model-00002-of-00006.bin",
521
+ "model.layers.11.self_attn.o_proj.weight": "pytorch_model-00002-of-00006.bin",
522
+ "model.layers.11.self_attn.q_proj.weight": "pytorch_model-00002-of-00006.bin",
523
+ "model.layers.11.self_attn.v_proj.weight": "pytorch_model-00002-of-00006.bin",
524
+ "model.layers.12.input_layernorm.weight": "pytorch_model-00002-of-00006.bin",
525
+ "model.layers.12.mlp.down_proj.weight": "pytorch_model-00002-of-00006.bin",
526
+ "model.layers.12.mlp.gate_proj.weight": "pytorch_model-00002-of-00006.bin",
527
+ "model.layers.12.mlp.up_proj.weight": "pytorch_model-00002-of-00006.bin",
528
+ "model.layers.12.post_attention_layernorm.weight": "pytorch_model-00002-of-00006.bin",
529
+ "model.layers.12.self_attn.k_proj.weight": "pytorch_model-00002-of-00006.bin",
530
+ "model.layers.12.self_attn.o_proj.weight": "pytorch_model-00002-of-00006.bin",
531
+ "model.layers.12.self_attn.q_proj.weight": "pytorch_model-00002-of-00006.bin",
532
+ "model.layers.12.self_attn.v_proj.weight": "pytorch_model-00002-of-00006.bin",
533
+ "model.layers.13.input_layernorm.weight": "pytorch_model-00002-of-00006.bin",
534
+ "model.layers.13.mlp.down_proj.weight": "pytorch_model-00002-of-00006.bin",
535
+ "model.layers.13.mlp.gate_proj.weight": "pytorch_model-00002-of-00006.bin",
536
+ "model.layers.13.mlp.up_proj.weight": "pytorch_model-00002-of-00006.bin",
537
+ "model.layers.13.post_attention_layernorm.weight": "pytorch_model-00002-of-00006.bin",
538
+ "model.layers.13.self_attn.k_proj.weight": "pytorch_model-00002-of-00006.bin",
539
+ "model.layers.13.self_attn.o_proj.weight": "pytorch_model-00002-of-00006.bin",
540
+ "model.layers.13.self_attn.q_proj.weight": "pytorch_model-00002-of-00006.bin",
541
+ "model.layers.13.self_attn.v_proj.weight": "pytorch_model-00002-of-00006.bin",
542
+ "model.layers.14.input_layernorm.weight": "pytorch_model-00002-of-00006.bin",
543
+ "model.layers.14.mlp.down_proj.weight": "pytorch_model-00002-of-00006.bin",
544
+ "model.layers.14.mlp.gate_proj.weight": "pytorch_model-00002-of-00006.bin",
545
+ "model.layers.14.mlp.up_proj.weight": "pytorch_model-00002-of-00006.bin",
546
+ "model.layers.14.post_attention_layernorm.weight": "pytorch_model-00002-of-00006.bin",
547
+ "model.layers.14.self_attn.k_proj.weight": "pytorch_model-00002-of-00006.bin",
548
+ "model.layers.14.self_attn.o_proj.weight": "pytorch_model-00002-of-00006.bin",
549
+ "model.layers.14.self_attn.q_proj.weight": "pytorch_model-00002-of-00006.bin",
550
+ "model.layers.14.self_attn.v_proj.weight": "pytorch_model-00002-of-00006.bin",
551
+ "model.layers.15.input_layernorm.weight": "pytorch_model-00003-of-00006.bin",
552
+ "model.layers.15.mlp.down_proj.weight": "pytorch_model-00003-of-00006.bin",
553
+ "model.layers.15.mlp.gate_proj.weight": "pytorch_model-00003-of-00006.bin",
554
+ "model.layers.15.mlp.up_proj.weight": "pytorch_model-00003-of-00006.bin",
555
+ "model.layers.15.post_attention_layernorm.weight": "pytorch_model-00003-of-00006.bin",
556
+ "model.layers.15.self_attn.k_proj.weight": "pytorch_model-00002-of-00006.bin",
557
+ "model.layers.15.self_attn.o_proj.weight": "pytorch_model-00003-of-00006.bin",
558
+ "model.layers.15.self_attn.q_proj.weight": "pytorch_model-00002-of-00006.bin",
559
+ "model.layers.15.self_attn.v_proj.weight": "pytorch_model-00003-of-00006.bin",
560
+ "model.layers.16.input_layernorm.weight": "pytorch_model-00003-of-00006.bin",
561
+ "model.layers.16.mlp.down_proj.weight": "pytorch_model-00003-of-00006.bin",
562
+ "model.layers.16.mlp.gate_proj.weight": "pytorch_model-00003-of-00006.bin",
563
+ "model.layers.16.mlp.up_proj.weight": "pytorch_model-00003-of-00006.bin",
564
+ "model.layers.16.post_attention_layernorm.weight": "pytorch_model-00003-of-00006.bin",
565
+ "model.layers.16.self_attn.k_proj.weight": "pytorch_model-00003-of-00006.bin",
566
+ "model.layers.16.self_attn.o_proj.weight": "pytorch_model-00003-of-00006.bin",
567
+ "model.layers.16.self_attn.q_proj.weight": "pytorch_model-00003-of-00006.bin",
568
+ "model.layers.16.self_attn.v_proj.weight": "pytorch_model-00003-of-00006.bin",
569
+ "model.layers.17.input_layernorm.weight": "pytorch_model-00003-of-00006.bin",
570
+ "model.layers.17.mlp.down_proj.weight": "pytorch_model-00003-of-00006.bin",
571
+ "model.layers.17.mlp.gate_proj.weight": "pytorch_model-00003-of-00006.bin",
572
+ "model.layers.17.mlp.up_proj.weight": "pytorch_model-00003-of-00006.bin",
573
+ "model.layers.17.post_attention_layernorm.weight": "pytorch_model-00003-of-00006.bin",
574
+ "model.layers.17.self_attn.k_proj.weight": "pytorch_model-00003-of-00006.bin",
575
+ "model.layers.17.self_attn.o_proj.weight": "pytorch_model-00003-of-00006.bin",
576
+ "model.layers.17.self_attn.q_proj.weight": "pytorch_model-00003-of-00006.bin",
577
+ "model.layers.17.self_attn.v_proj.weight": "pytorch_model-00003-of-00006.bin",
578
+ "model.layers.18.input_layernorm.weight": "pytorch_model-00003-of-00006.bin",
579
+ "model.layers.18.mlp.down_proj.weight": "pytorch_model-00003-of-00006.bin",
580
+ "model.layers.18.mlp.gate_proj.weight": "pytorch_model-00003-of-00006.bin",
581
+ "model.layers.18.mlp.up_proj.weight": "pytorch_model-00003-of-00006.bin",
582
+ "model.layers.18.post_attention_layernorm.weight": "pytorch_model-00003-of-00006.bin",
583
+ "model.layers.18.self_attn.k_proj.weight": "pytorch_model-00003-of-00006.bin",
584
+ "model.layers.18.self_attn.o_proj.weight": "pytorch_model-00003-of-00006.bin",
585
+ "model.layers.18.self_attn.q_proj.weight": "pytorch_model-00003-of-00006.bin",
586
+ "model.layers.18.self_attn.v_proj.weight": "pytorch_model-00003-of-00006.bin",
587
+ "model.layers.19.input_layernorm.weight": "pytorch_model-00003-of-00006.bin",
588
+ "model.layers.19.mlp.down_proj.weight": "pytorch_model-00003-of-00006.bin",
589
+ "model.layers.19.mlp.gate_proj.weight": "pytorch_model-00003-of-00006.bin",
590
+ "model.layers.19.mlp.up_proj.weight": "pytorch_model-00003-of-00006.bin",
591
+ "model.layers.19.post_attention_layernorm.weight": "pytorch_model-00003-of-00006.bin",
592
+ "model.layers.19.self_attn.k_proj.weight": "pytorch_model-00003-of-00006.bin",
593
+ "model.layers.19.self_attn.o_proj.weight": "pytorch_model-00003-of-00006.bin",
594
+ "model.layers.19.self_attn.q_proj.weight": "pytorch_model-00003-of-00006.bin",
595
+ "model.layers.19.self_attn.v_proj.weight": "pytorch_model-00003-of-00006.bin",
596
+ "model.layers.2.input_layernorm.weight": "pytorch_model-00001-of-00006.bin",
597
+ "model.layers.2.mlp.down_proj.weight": "pytorch_model-00001-of-00006.bin",
598
+ "model.layers.2.mlp.gate_proj.weight": "pytorch_model-00001-of-00006.bin",
599
+ "model.layers.2.mlp.up_proj.weight": "pytorch_model-00001-of-00006.bin",
600
+ "model.layers.2.post_attention_layernorm.weight": "pytorch_model-00001-of-00006.bin",
601
+ "model.layers.2.self_attn.k_proj.weight": "pytorch_model-00001-of-00006.bin",
602
+ "model.layers.2.self_attn.o_proj.weight": "pytorch_model-00001-of-00006.bin",
603
+ "model.layers.2.self_attn.q_proj.weight": "pytorch_model-00001-of-00006.bin",
604
+ "model.layers.2.self_attn.v_proj.weight": "pytorch_model-00001-of-00006.bin",
605
+ "model.layers.20.input_layernorm.weight": "pytorch_model-00003-of-00006.bin",
606
+ "model.layers.20.mlp.down_proj.weight": "pytorch_model-00003-of-00006.bin",
607
+ "model.layers.20.mlp.gate_proj.weight": "pytorch_model-00003-of-00006.bin",
608
+ "model.layers.20.mlp.up_proj.weight": "pytorch_model-00003-of-00006.bin",
609
+ "model.layers.20.post_attention_layernorm.weight": "pytorch_model-00003-of-00006.bin",
610
+ "model.layers.20.self_attn.k_proj.weight": "pytorch_model-00003-of-00006.bin",
611
+ "model.layers.20.self_attn.o_proj.weight": "pytorch_model-00003-of-00006.bin",
612
+ "model.layers.20.self_attn.q_proj.weight": "pytorch_model-00003-of-00006.bin",
613
+ "model.layers.20.self_attn.v_proj.weight": "pytorch_model-00003-of-00006.bin",
614
+ "model.layers.21.input_layernorm.weight": "pytorch_model-00003-of-00006.bin",
615
+ "model.layers.21.mlp.down_proj.weight": "pytorch_model-00003-of-00006.bin",
616
+ "model.layers.21.mlp.gate_proj.weight": "pytorch_model-00003-of-00006.bin",
617
+ "model.layers.21.mlp.up_proj.weight": "pytorch_model-00003-of-00006.bin",
618
+ "model.layers.21.post_attention_layernorm.weight": "pytorch_model-00003-of-00006.bin",
619
+ "model.layers.21.self_attn.k_proj.weight": "pytorch_model-00003-of-00006.bin",
620
+ "model.layers.21.self_attn.o_proj.weight": "pytorch_model-00003-of-00006.bin",
621
+ "model.layers.21.self_attn.q_proj.weight": "pytorch_model-00003-of-00006.bin",
622
+ "model.layers.21.self_attn.v_proj.weight": "pytorch_model-00003-of-00006.bin",
623
+ "model.layers.22.input_layernorm.weight": "pytorch_model-00003-of-00006.bin",
624
+ "model.layers.22.mlp.down_proj.weight": "pytorch_model-00003-of-00006.bin",
625
+ "model.layers.22.mlp.gate_proj.weight": "pytorch_model-00003-of-00006.bin",
626
+ "model.layers.22.mlp.up_proj.weight": "pytorch_model-00003-of-00006.bin",
627
+ "model.layers.22.post_attention_layernorm.weight": "pytorch_model-00003-of-00006.bin",
628
+ "model.layers.22.self_attn.k_proj.weight": "pytorch_model-00003-of-00006.bin",
629
+ "model.layers.22.self_attn.o_proj.weight": "pytorch_model-00003-of-00006.bin",
630
+ "model.layers.22.self_attn.q_proj.weight": "pytorch_model-00003-of-00006.bin",
631
+ "model.layers.22.self_attn.v_proj.weight": "pytorch_model-00003-of-00006.bin",
632
+ "model.layers.23.input_layernorm.weight": "pytorch_model-00004-of-00006.bin",
633
+ "model.layers.23.mlp.down_proj.weight": "pytorch_model-00004-of-00006.bin",
634
+ "model.layers.23.mlp.gate_proj.weight": "pytorch_model-00004-of-00006.bin",
635
+ "model.layers.23.mlp.up_proj.weight": "pytorch_model-00004-of-00006.bin",
636
+ "model.layers.23.post_attention_layernorm.weight": "pytorch_model-00004-of-00006.bin",
637
+ "model.layers.23.self_attn.k_proj.weight": "pytorch_model-00004-of-00006.bin",
638
+ "model.layers.23.self_attn.o_proj.weight": "pytorch_model-00004-of-00006.bin",
639
+ "model.layers.23.self_attn.q_proj.weight": "pytorch_model-00004-of-00006.bin",
640
+ "model.layers.23.self_attn.v_proj.weight": "pytorch_model-00004-of-00006.bin",
641
+ "model.layers.24.input_layernorm.weight": "pytorch_model-00004-of-00006.bin",
642
+ "model.layers.24.mlp.down_proj.weight": "pytorch_model-00004-of-00006.bin",
643
+ "model.layers.24.mlp.gate_proj.weight": "pytorch_model-00004-of-00006.bin",
644
+ "model.layers.24.mlp.up_proj.weight": "pytorch_model-00004-of-00006.bin",
645
+ "model.layers.24.post_attention_layernorm.weight": "pytorch_model-00004-of-00006.bin",
646
+ "model.layers.24.self_attn.k_proj.weight": "pytorch_model-00004-of-00006.bin",
647
+ "model.layers.24.self_attn.o_proj.weight": "pytorch_model-00004-of-00006.bin",
648
+ "model.layers.24.self_attn.q_proj.weight": "pytorch_model-00004-of-00006.bin",
649
+ "model.layers.24.self_attn.v_proj.weight": "pytorch_model-00004-of-00006.bin",
650
+ "model.layers.25.input_layernorm.weight": "pytorch_model-00004-of-00006.bin",
651
+ "model.layers.25.mlp.down_proj.weight": "pytorch_model-00004-of-00006.bin",
652
+ "model.layers.25.mlp.gate_proj.weight": "pytorch_model-00004-of-00006.bin",
653
+ "model.layers.25.mlp.up_proj.weight": "pytorch_model-00004-of-00006.bin",
654
+ "model.layers.25.post_attention_layernorm.weight": "pytorch_model-00004-of-00006.bin",
655
+ "model.layers.25.self_attn.k_proj.weight": "pytorch_model-00004-of-00006.bin",
656
+ "model.layers.25.self_attn.o_proj.weight": "pytorch_model-00004-of-00006.bin",
657
+ "model.layers.25.self_attn.q_proj.weight": "pytorch_model-00004-of-00006.bin",
658
+ "model.layers.25.self_attn.v_proj.weight": "pytorch_model-00004-of-00006.bin",
659
+ "model.layers.26.input_layernorm.weight": "pytorch_model-00004-of-00006.bin",
660
+ "model.layers.26.mlp.down_proj.weight": "pytorch_model-00004-of-00006.bin",
661
+ "model.layers.26.mlp.gate_proj.weight": "pytorch_model-00004-of-00006.bin",
662
+ "model.layers.26.mlp.up_proj.weight": "pytorch_model-00004-of-00006.bin",
663
+ "model.layers.26.post_attention_layernorm.weight": "pytorch_model-00004-of-00006.bin",
664
+ "model.layers.26.self_attn.k_proj.weight": "pytorch_model-00004-of-00006.bin",
665
+ "model.layers.26.self_attn.o_proj.weight": "pytorch_model-00004-of-00006.bin",
666
+ "model.layers.26.self_attn.q_proj.weight": "pytorch_model-00004-of-00006.bin",
667
+ "model.layers.26.self_attn.v_proj.weight": "pytorch_model-00004-of-00006.bin",
668
+ "model.layers.27.input_layernorm.weight": "pytorch_model-00004-of-00006.bin",
669
+ "model.layers.27.mlp.down_proj.weight": "pytorch_model-00004-of-00006.bin",
670
+ "model.layers.27.mlp.gate_proj.weight": "pytorch_model-00004-of-00006.bin",
671
+ "model.layers.27.mlp.up_proj.weight": "pytorch_model-00004-of-00006.bin",
672
+ "model.layers.27.post_attention_layernorm.weight": "pytorch_model-00004-of-00006.bin",
673
+ "model.layers.27.self_attn.k_proj.weight": "pytorch_model-00004-of-00006.bin",
674
+ "model.layers.27.self_attn.o_proj.weight": "pytorch_model-00004-of-00006.bin",
675
+ "model.layers.27.self_attn.q_proj.weight": "pytorch_model-00004-of-00006.bin",
676
+ "model.layers.27.self_attn.v_proj.weight": "pytorch_model-00004-of-00006.bin",
677
+ "model.layers.28.input_layernorm.weight": "pytorch_model-00004-of-00006.bin",
678
+ "model.layers.28.mlp.down_proj.weight": "pytorch_model-00004-of-00006.bin",
679
+ "model.layers.28.mlp.gate_proj.weight": "pytorch_model-00004-of-00006.bin",
680
+ "model.layers.28.mlp.up_proj.weight": "pytorch_model-00004-of-00006.bin",
681
+ "model.layers.28.post_attention_layernorm.weight": "pytorch_model-00004-of-00006.bin",
682
+ "model.layers.28.self_attn.k_proj.weight": "pytorch_model-00004-of-00006.bin",
683
+ "model.layers.28.self_attn.o_proj.weight": "pytorch_model-00004-of-00006.bin",
684
+ "model.layers.28.self_attn.q_proj.weight": "pytorch_model-00004-of-00006.bin",
685
+ "model.layers.28.self_attn.v_proj.weight": "pytorch_model-00004-of-00006.bin",
686
+ "model.layers.29.input_layernorm.weight": "pytorch_model-00004-of-00006.bin",
687
+ "model.layers.29.mlp.down_proj.weight": "pytorch_model-00004-of-00006.bin",
688
+ "model.layers.29.mlp.gate_proj.weight": "pytorch_model-00004-of-00006.bin",
689
+ "model.layers.29.mlp.up_proj.weight": "pytorch_model-00004-of-00006.bin",
690
+ "model.layers.29.post_attention_layernorm.weight": "pytorch_model-00004-of-00006.bin",
691
+ "model.layers.29.self_attn.k_proj.weight": "pytorch_model-00004-of-00006.bin",
692
+ "model.layers.29.self_attn.o_proj.weight": "pytorch_model-00004-of-00006.bin",
693
+ "model.layers.29.self_attn.q_proj.weight": "pytorch_model-00004-of-00006.bin",
694
+ "model.layers.29.self_attn.v_proj.weight": "pytorch_model-00004-of-00006.bin",
695
+ "model.layers.3.input_layernorm.weight": "pytorch_model-00001-of-00006.bin",
696
+ "model.layers.3.mlp.down_proj.weight": "pytorch_model-00001-of-00006.bin",
697
+ "model.layers.3.mlp.gate_proj.weight": "pytorch_model-00001-of-00006.bin",
698
+ "model.layers.3.mlp.up_proj.weight": "pytorch_model-00001-of-00006.bin",
699
+ "model.layers.3.post_attention_layernorm.weight": "pytorch_model-00001-of-00006.bin",
700
+ "model.layers.3.self_attn.k_proj.weight": "pytorch_model-00001-of-00006.bin",
701
+ "model.layers.3.self_attn.o_proj.weight": "pytorch_model-00001-of-00006.bin",
702
+ "model.layers.3.self_attn.q_proj.weight": "pytorch_model-00001-of-00006.bin",
703
+ "model.layers.3.self_attn.v_proj.weight": "pytorch_model-00001-of-00006.bin",
704
+ "model.layers.30.input_layernorm.weight": "pytorch_model-00005-of-00006.bin",
705
+ "model.layers.30.mlp.down_proj.weight": "pytorch_model-00005-of-00006.bin",
706
+ "model.layers.30.mlp.gate_proj.weight": "pytorch_model-00004-of-00006.bin",
707
+ "model.layers.30.mlp.up_proj.weight": "pytorch_model-00004-of-00006.bin",
708
+ "model.layers.30.post_attention_layernorm.weight": "pytorch_model-00005-of-00006.bin",
709
+ "model.layers.30.self_attn.k_proj.weight": "pytorch_model-00004-of-00006.bin",
710
+ "model.layers.30.self_attn.o_proj.weight": "pytorch_model-00004-of-00006.bin",
711
+ "model.layers.30.self_attn.q_proj.weight": "pytorch_model-00004-of-00006.bin",
712
+ "model.layers.30.self_attn.v_proj.weight": "pytorch_model-00004-of-00006.bin",
713
+ "model.layers.31.input_layernorm.weight": "pytorch_model-00005-of-00006.bin",
714
+ "model.layers.31.mlp.down_proj.weight": "pytorch_model-00005-of-00006.bin",
715
+ "model.layers.31.mlp.gate_proj.weight": "pytorch_model-00005-of-00006.bin",
716
+ "model.layers.31.mlp.up_proj.weight": "pytorch_model-00005-of-00006.bin",
717
+ "model.layers.31.post_attention_layernorm.weight": "pytorch_model-00005-of-00006.bin",
718
+ "model.layers.31.self_attn.k_proj.weight": "pytorch_model-00005-of-00006.bin",
719
+ "model.layers.31.self_attn.o_proj.weight": "pytorch_model-00005-of-00006.bin",
720
+ "model.layers.31.self_attn.q_proj.weight": "pytorch_model-00005-of-00006.bin",
721
+ "model.layers.31.self_attn.v_proj.weight": "pytorch_model-00005-of-00006.bin",
722
+ "model.layers.32.input_layernorm.weight": "pytorch_model-00005-of-00006.bin",
723
+ "model.layers.32.mlp.down_proj.weight": "pytorch_model-00005-of-00006.bin",
724
+ "model.layers.32.mlp.gate_proj.weight": "pytorch_model-00005-of-00006.bin",
725
+ "model.layers.32.mlp.up_proj.weight": "pytorch_model-00005-of-00006.bin",
726
+ "model.layers.32.post_attention_layernorm.weight": "pytorch_model-00005-of-00006.bin",
727
+ "model.layers.32.self_attn.k_proj.weight": "pytorch_model-00005-of-00006.bin",
728
+ "model.layers.32.self_attn.o_proj.weight": "pytorch_model-00005-of-00006.bin",
729
+ "model.layers.32.self_attn.q_proj.weight": "pytorch_model-00005-of-00006.bin",
730
+ "model.layers.32.self_attn.v_proj.weight": "pytorch_model-00005-of-00006.bin",
731
+ "model.layers.33.input_layernorm.weight": "pytorch_model-00005-of-00006.bin",
732
+ "model.layers.33.mlp.down_proj.weight": "pytorch_model-00005-of-00006.bin",
733
+ "model.layers.33.mlp.gate_proj.weight": "pytorch_model-00005-of-00006.bin",
734
+ "model.layers.33.mlp.up_proj.weight": "pytorch_model-00005-of-00006.bin",
735
+ "model.layers.33.post_attention_layernorm.weight": "pytorch_model-00005-of-00006.bin",
736
+ "model.layers.33.self_attn.k_proj.weight": "pytorch_model-00005-of-00006.bin",
737
+ "model.layers.33.self_attn.o_proj.weight": "pytorch_model-00005-of-00006.bin",
738
+ "model.layers.33.self_attn.q_proj.weight": "pytorch_model-00005-of-00006.bin",
739
+ "model.layers.33.self_attn.v_proj.weight": "pytorch_model-00005-of-00006.bin",
740
+ "model.layers.34.input_layernorm.weight": "pytorch_model-00005-of-00006.bin",
741
+ "model.layers.34.mlp.down_proj.weight": "pytorch_model-00005-of-00006.bin",
742
+ "model.layers.34.mlp.gate_proj.weight": "pytorch_model-00005-of-00006.bin",
743
+ "model.layers.34.mlp.up_proj.weight": "pytorch_model-00005-of-00006.bin",
744
+ "model.layers.34.post_attention_layernorm.weight": "pytorch_model-00005-of-00006.bin",
745
+ "model.layers.34.self_attn.k_proj.weight": "pytorch_model-00005-of-00006.bin",
746
+ "model.layers.34.self_attn.o_proj.weight": "pytorch_model-00005-of-00006.bin",
747
+ "model.layers.34.self_attn.q_proj.weight": "pytorch_model-00005-of-00006.bin",
748
+ "model.layers.34.self_attn.v_proj.weight": "pytorch_model-00005-of-00006.bin",
749
+ "model.layers.35.input_layernorm.weight": "pytorch_model-00005-of-00006.bin",
750
+ "model.layers.35.mlp.down_proj.weight": "pytorch_model-00005-of-00006.bin",
751
+ "model.layers.35.mlp.gate_proj.weight": "pytorch_model-00005-of-00006.bin",
752
+ "model.layers.35.mlp.up_proj.weight": "pytorch_model-00005-of-00006.bin",
753
+ "model.layers.35.post_attention_layernorm.weight": "pytorch_model-00005-of-00006.bin",
754
+ "model.layers.35.self_attn.k_proj.weight": "pytorch_model-00005-of-00006.bin",
755
+ "model.layers.35.self_attn.o_proj.weight": "pytorch_model-00005-of-00006.bin",
756
+ "model.layers.35.self_attn.q_proj.weight": "pytorch_model-00005-of-00006.bin",
757
+ "model.layers.35.self_attn.v_proj.weight": "pytorch_model-00005-of-00006.bin",
758
+ "model.layers.36.input_layernorm.weight": "pytorch_model-00005-of-00006.bin",
759
+ "model.layers.36.mlp.down_proj.weight": "pytorch_model-00005-of-00006.bin",
760
+ "model.layers.36.mlp.gate_proj.weight": "pytorch_model-00005-of-00006.bin",
761
+ "model.layers.36.mlp.up_proj.weight": "pytorch_model-00005-of-00006.bin",
762
+ "model.layers.36.post_attention_layernorm.weight": "pytorch_model-00005-of-00006.bin",
763
+ "model.layers.36.self_attn.k_proj.weight": "pytorch_model-00005-of-00006.bin",
764
+ "model.layers.36.self_attn.o_proj.weight": "pytorch_model-00005-of-00006.bin",
765
+ "model.layers.36.self_attn.q_proj.weight": "pytorch_model-00005-of-00006.bin",
766
+ "model.layers.36.self_attn.v_proj.weight": "pytorch_model-00005-of-00006.bin",
767
+ "model.layers.37.input_layernorm.weight": "pytorch_model-00005-of-00006.bin",
768
+ "model.layers.37.mlp.down_proj.weight": "pytorch_model-00005-of-00006.bin",
769
+ "model.layers.37.mlp.gate_proj.weight": "pytorch_model-00005-of-00006.bin",
770
+ "model.layers.37.mlp.up_proj.weight": "pytorch_model-00005-of-00006.bin",
771
+ "model.layers.37.post_attention_layernorm.weight": "pytorch_model-00005-of-00006.bin",
772
+ "model.layers.37.self_attn.k_proj.weight": "pytorch_model-00005-of-00006.bin",
773
+ "model.layers.37.self_attn.o_proj.weight": "pytorch_model-00005-of-00006.bin",
774
+ "model.layers.37.self_attn.q_proj.weight": "pytorch_model-00005-of-00006.bin",
775
+ "model.layers.37.self_attn.v_proj.weight": "pytorch_model-00005-of-00006.bin",
776
+ "model.layers.38.input_layernorm.weight": "pytorch_model-00006-of-00006.bin",
777
+ "model.layers.38.mlp.down_proj.weight": "pytorch_model-00006-of-00006.bin",
778
+ "model.layers.38.mlp.gate_proj.weight": "pytorch_model-00005-of-00006.bin",
779
+ "model.layers.38.mlp.up_proj.weight": "pytorch_model-00006-of-00006.bin",
780
+ "model.layers.38.post_attention_layernorm.weight": "pytorch_model-00006-of-00006.bin",
781
+ "model.layers.38.self_attn.k_proj.weight": "pytorch_model-00005-of-00006.bin",
782
+ "model.layers.38.self_attn.o_proj.weight": "pytorch_model-00005-of-00006.bin",
783
+ "model.layers.38.self_attn.q_proj.weight": "pytorch_model-00005-of-00006.bin",
784
+ "model.layers.38.self_attn.v_proj.weight": "pytorch_model-00005-of-00006.bin",
785
+ "model.layers.39.input_layernorm.weight": "pytorch_model-00006-of-00006.bin",
786
+ "model.layers.39.mlp.down_proj.weight": "pytorch_model-00006-of-00006.bin",
787
+ "model.layers.39.mlp.gate_proj.weight": "pytorch_model-00006-of-00006.bin",
788
+ "model.layers.39.mlp.up_proj.weight": "pytorch_model-00006-of-00006.bin",
789
+ "model.layers.39.post_attention_layernorm.weight": "pytorch_model-00006-of-00006.bin",
790
+ "model.layers.39.self_attn.k_proj.weight": "pytorch_model-00006-of-00006.bin",
791
+ "model.layers.39.self_attn.o_proj.weight": "pytorch_model-00006-of-00006.bin",
792
+ "model.layers.39.self_attn.q_proj.weight": "pytorch_model-00006-of-00006.bin",
793
+ "model.layers.39.self_attn.v_proj.weight": "pytorch_model-00006-of-00006.bin",
794
+ "model.layers.4.input_layernorm.weight": "pytorch_model-00001-of-00006.bin",
795
+ "model.layers.4.mlp.down_proj.weight": "pytorch_model-00001-of-00006.bin",
796
+ "model.layers.4.mlp.gate_proj.weight": "pytorch_model-00001-of-00006.bin",
797
+ "model.layers.4.mlp.up_proj.weight": "pytorch_model-00001-of-00006.bin",
798
+ "model.layers.4.post_attention_layernorm.weight": "pytorch_model-00001-of-00006.bin",
799
+ "model.layers.4.self_attn.k_proj.weight": "pytorch_model-00001-of-00006.bin",
800
+ "model.layers.4.self_attn.o_proj.weight": "pytorch_model-00001-of-00006.bin",
801
+ "model.layers.4.self_attn.q_proj.weight": "pytorch_model-00001-of-00006.bin",
802
+ "model.layers.4.self_attn.v_proj.weight": "pytorch_model-00001-of-00006.bin",
803
+ "model.layers.5.input_layernorm.weight": "pytorch_model-00001-of-00006.bin",
804
+ "model.layers.5.mlp.down_proj.weight": "pytorch_model-00001-of-00006.bin",
805
+ "model.layers.5.mlp.gate_proj.weight": "pytorch_model-00001-of-00006.bin",
806
+ "model.layers.5.mlp.up_proj.weight": "pytorch_model-00001-of-00006.bin",
807
+ "model.layers.5.post_attention_layernorm.weight": "pytorch_model-00001-of-00006.bin",
808
+ "model.layers.5.self_attn.k_proj.weight": "pytorch_model-00001-of-00006.bin",
809
+ "model.layers.5.self_attn.o_proj.weight": "pytorch_model-00001-of-00006.bin",
810
+ "model.layers.5.self_attn.q_proj.weight": "pytorch_model-00001-of-00006.bin",
811
+ "model.layers.5.self_attn.v_proj.weight": "pytorch_model-00001-of-00006.bin",
812
+ "model.layers.6.input_layernorm.weight": "pytorch_model-00001-of-00006.bin",
813
+ "model.layers.6.mlp.down_proj.weight": "pytorch_model-00001-of-00006.bin",
814
+ "model.layers.6.mlp.gate_proj.weight": "pytorch_model-00001-of-00006.bin",
815
+ "model.layers.6.mlp.up_proj.weight": "pytorch_model-00001-of-00006.bin",
816
+ "model.layers.6.post_attention_layernorm.weight": "pytorch_model-00001-of-00006.bin",
817
+ "model.layers.6.self_attn.k_proj.weight": "pytorch_model-00001-of-00006.bin",
818
+ "model.layers.6.self_attn.o_proj.weight": "pytorch_model-00001-of-00006.bin",
819
+ "model.layers.6.self_attn.q_proj.weight": "pytorch_model-00001-of-00006.bin",
820
+ "model.layers.6.self_attn.v_proj.weight": "pytorch_model-00001-of-00006.bin",
821
+ "model.layers.7.input_layernorm.weight": "pytorch_model-00002-of-00006.bin",
822
+ "model.layers.7.mlp.down_proj.weight": "pytorch_model-00002-of-00006.bin",
823
+ "model.layers.7.mlp.gate_proj.weight": "pytorch_model-00002-of-00006.bin",
824
+ "model.layers.7.mlp.up_proj.weight": "pytorch_model-00002-of-00006.bin",
825
+ "model.layers.7.post_attention_layernorm.weight": "pytorch_model-00002-of-00006.bin",
826
+ "model.layers.7.self_attn.k_proj.weight": "pytorch_model-00001-of-00006.bin",
827
+ "model.layers.7.self_attn.o_proj.weight": "pytorch_model-00001-of-00006.bin",
828
+ "model.layers.7.self_attn.q_proj.weight": "pytorch_model-00001-of-00006.bin",
829
+ "model.layers.7.self_attn.v_proj.weight": "pytorch_model-00001-of-00006.bin",
830
+ "model.layers.8.input_layernorm.weight": "pytorch_model-00002-of-00006.bin",
831
+ "model.layers.8.mlp.down_proj.weight": "pytorch_model-00002-of-00006.bin",
832
+ "model.layers.8.mlp.gate_proj.weight": "pytorch_model-00002-of-00006.bin",
833
+ "model.layers.8.mlp.up_proj.weight": "pytorch_model-00002-of-00006.bin",
834
+ "model.layers.8.post_attention_layernorm.weight": "pytorch_model-00002-of-00006.bin",
835
+ "model.layers.8.self_attn.k_proj.weight": "pytorch_model-00002-of-00006.bin",
836
+ "model.layers.8.self_attn.o_proj.weight": "pytorch_model-00002-of-00006.bin",
837
+ "model.layers.8.self_attn.q_proj.weight": "pytorch_model-00002-of-00006.bin",
838
+ "model.layers.8.self_attn.v_proj.weight": "pytorch_model-00002-of-00006.bin",
839
+ "model.layers.9.input_layernorm.weight": "pytorch_model-00002-of-00006.bin",
840
+ "model.layers.9.mlp.down_proj.weight": "pytorch_model-00002-of-00006.bin",
841
+ "model.layers.9.mlp.gate_proj.weight": "pytorch_model-00002-of-00006.bin",
842
+ "model.layers.9.mlp.up_proj.weight": "pytorch_model-00002-of-00006.bin",
843
+ "model.layers.9.post_attention_layernorm.weight": "pytorch_model-00002-of-00006.bin",
844
+ "model.layers.9.self_attn.k_proj.weight": "pytorch_model-00002-of-00006.bin",
845
+ "model.layers.9.self_attn.o_proj.weight": "pytorch_model-00002-of-00006.bin",
846
+ "model.layers.9.self_attn.q_proj.weight": "pytorch_model-00002-of-00006.bin",
847
+ "model.layers.9.self_attn.v_proj.weight": "pytorch_model-00002-of-00006.bin",
848
+ "model.norm.weight": "pytorch_model-00006-of-00006.bin"
849
+ }
850
+ }