Andreas Koepf commited on
Commit
0319e91
1 Parent(s): d9f2927

pad embeddings to multiple of 128

Browse files
config.json CHANGED
@@ -20,7 +20,7 @@
20
  "rope_scaling": null,
21
  "tie_word_embeddings": false,
22
  "torch_dtype": "bfloat16",
23
- "transformers_version": "4.31.0",
24
  "use_cache": true,
25
  "vocab_size": 32007
26
  }
 
20
  "rope_scaling": null,
21
  "tie_word_embeddings": false,
22
  "torch_dtype": "bfloat16",
23
+ "transformers_version": "4.32.1",
24
  "use_cache": true,
25
  "vocab_size": 32007
26
  }
generation_config.json CHANGED
@@ -3,5 +3,5 @@
3
  "bos_token_id": 32005,
4
  "eos_token_id": 32006,
5
  "pad_token_id": 0,
6
- "transformers_version": "4.31.0"
7
  }
 
3
  "bos_token_id": 32005,
4
  "eos_token_id": 32006,
5
  "pad_token_id": 0,
6
+ "transformers_version": "4.32.1"
7
  }
pytorch_model-00001-of-00015.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:79e727d96325e58363249ec93e02202f59036d6b68f796c2a049098875a62652
3
- size 9852720092
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5693d282af8c41a185c93fc2161e09abb4b5ad39fe9b6328ec55880b0f3ac693
3
+ size 9854698469
pytorch_model-00002-of-00015.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:5c612ad81c4448e8dcb6cf27ea517da691dcd9cb024db72c45540e68abbaee4f
3
- size 9798112992
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7b3276da1edcf36f8bd07bd8bfbb285a5ce273f18576ce80ce1e610b79e870b7
3
+ size 9798108893
pytorch_model-00003-of-00015.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:6909ddaf1a83177d1306e33312075c479c26f5f62b2a5753bb1da120228cdeb4
3
- size 9965883644
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:997faaee1e4f721f3b4741a0a239057d07bfb2a77fd117be93e40a82f5132434
3
+ size 9965880015
pytorch_model-00004-of-00015.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:9708c7a22321bcbdc8d5d6c92de0d9699309d50b20b1cd723e6759525ab0a7b7
3
- size 9798079504
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:237582071bd98134d4c1ace75fc7086be4e1920fd35187c3a887c8a0a84f4c72
3
+ size 9798075429
pytorch_model-00005-of-00015.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:0279705ab247d84ca980d60bc185be693451164096eb25dd46c2aa3501846ddd
3
- size 9798113032
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:12a42ef611c4257826e62f78f68fcc600488f706c036ec03bf6bb992c779207d
3
+ size 9798108873
pytorch_model-00006-of-00015.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:ec3ac6a0d1cc28f8f8ef33fc3dc8bf1c762f21763cd2defc8ae5d6d19f621b86
3
- size 9798113056
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:209f4136bb1231c7529a479035f64f21af1722c425cfb37cea8324031388a8e9
3
+ size 9798108893
pytorch_model-00007-of-00015.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:93bbec5e1c205bde225fe25fc678b4a767ae4592c1578f6b1f8a9139e7eb6b01
3
- size 9965883644
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e0ff3dc22f681c9ff60eb0cd8cfbbd871b3abaecbe03d368813dbe069b549150
3
+ size 9965880015
pytorch_model-00008-of-00015.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:f59f0c15338986d18569a8113991170cb360d30de2d649cbe69f41582ee6ee52
3
- size 9798079504
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:109968c469b0558aac423cbf15b286cfbf3a08fed0e6bf840f4ec15428b43f4b
3
+ size 9798075429
pytorch_model-00009-of-00015.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:13dd585ddd5bfbe3ea6308bff39adaaa5d6cc61a2b46864d5adf079977dc4c5f
3
- size 9798113032
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0921cba072f174d91ba82206899492a2b3b677068dcef36d891aeabbbb016cf0
3
+ size 9798108873
pytorch_model-00010-of-00015.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:d91a65d3e10fb476c765df6c4fc159b7a547032b96c01fa96d6b8075dea763c3
3
- size 9798113056
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:485ab12fdbdcdb14af91b39c337a0f12dc7d96dfe93d15b933c1c411a541c46c
3
+ size 9798108893
pytorch_model-00011-of-00015.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:0d1077ff350755170569856b6e181ed53ccb28d87e28e57f695beb75973b7d4c
3
- size 9965883644
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b2fbbed4fbf68dbe0d826f85ed49f15d27aab4932657b6d7b1d593a4df170a9c
3
+ size 9965880015
pytorch_model-00012-of-00015.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:fe094728ef37207df7ef2b6a7efd5c2a11081d77657fa9b54f4299b002fb273e
3
- size 9798079504
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f86559f4aa8330ada47d5d0bd9c6c04cb3ce8675b545a97ea839bc95e05bb441
3
+ size 9798075429
pytorch_model-00013-of-00015.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:1a4cdf566d45c2d2f94ca94cbfd09007ef32aa7cd54d9157cb36a095adb18618
3
- size 9798113032
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7cf245007d308d49fecb28dd5c5e765a486a6cfa24d0c957fa597fc2aa2bc3b5
3
+ size 9798108873
pytorch_model-00014-of-00015.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:77ad4b712582397442685feef902c5e5331c2fe625c8079ff4a1595026a1aa17
3
- size 9496137832
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:04d518cd19ec244c0b93dc391c4654281dff721e05f0d6c19a364fedcd0094c7
3
+ size 9496134339
pytorch_model-00015-of-00015.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:656389099e94c6629ddf790f7e997bf26cc18f5eeba19f58ad7e30c29f28e81e
3
- size 524404101
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:70e19f991951f004336ae9af2f9db43ebd901ba1849104b0d58f77e4f213e549
3
+ size 526386090
pytorch_model.bin.index.json CHANGED
@@ -1,6 +1,6 @@
1
  {
2
  "metadata": {
3
- "total_size": 137953546240
4
  },
5
  "weight_map": {
6
  "lm_head.weight": "pytorch_model-00015-of-00015.bin",
@@ -13,7 +13,6 @@
13
  "model.layers.0.self_attn.k_proj.weight": "pytorch_model-00001-of-00015.bin",
14
  "model.layers.0.self_attn.o_proj.weight": "pytorch_model-00001-of-00015.bin",
15
  "model.layers.0.self_attn.q_proj.weight": "pytorch_model-00001-of-00015.bin",
16
- "model.layers.0.self_attn.rotary_emb.inv_freq": "pytorch_model-00001-of-00015.bin",
17
  "model.layers.0.self_attn.v_proj.weight": "pytorch_model-00001-of-00015.bin",
18
  "model.layers.1.input_layernorm.weight": "pytorch_model-00001-of-00015.bin",
19
  "model.layers.1.mlp.down_proj.weight": "pytorch_model-00001-of-00015.bin",
@@ -23,7 +22,6 @@
23
  "model.layers.1.self_attn.k_proj.weight": "pytorch_model-00001-of-00015.bin",
24
  "model.layers.1.self_attn.o_proj.weight": "pytorch_model-00001-of-00015.bin",
25
  "model.layers.1.self_attn.q_proj.weight": "pytorch_model-00001-of-00015.bin",
26
- "model.layers.1.self_attn.rotary_emb.inv_freq": "pytorch_model-00001-of-00015.bin",
27
  "model.layers.1.self_attn.v_proj.weight": "pytorch_model-00001-of-00015.bin",
28
  "model.layers.10.input_layernorm.weight": "pytorch_model-00002-of-00015.bin",
29
  "model.layers.10.mlp.down_proj.weight": "pytorch_model-00002-of-00015.bin",
@@ -33,7 +31,6 @@
33
  "model.layers.10.self_attn.k_proj.weight": "pytorch_model-00002-of-00015.bin",
34
  "model.layers.10.self_attn.o_proj.weight": "pytorch_model-00002-of-00015.bin",
35
  "model.layers.10.self_attn.q_proj.weight": "pytorch_model-00002-of-00015.bin",
36
- "model.layers.10.self_attn.rotary_emb.inv_freq": "pytorch_model-00002-of-00015.bin",
37
  "model.layers.10.self_attn.v_proj.weight": "pytorch_model-00002-of-00015.bin",
38
  "model.layers.11.input_layernorm.weight": "pytorch_model-00003-of-00015.bin",
39
  "model.layers.11.mlp.down_proj.weight": "pytorch_model-00003-of-00015.bin",
@@ -43,7 +40,6 @@
43
  "model.layers.11.self_attn.k_proj.weight": "pytorch_model-00002-of-00015.bin",
44
  "model.layers.11.self_attn.o_proj.weight": "pytorch_model-00002-of-00015.bin",
45
  "model.layers.11.self_attn.q_proj.weight": "pytorch_model-00002-of-00015.bin",
46
- "model.layers.11.self_attn.rotary_emb.inv_freq": "pytorch_model-00002-of-00015.bin",
47
  "model.layers.11.self_attn.v_proj.weight": "pytorch_model-00002-of-00015.bin",
48
  "model.layers.12.input_layernorm.weight": "pytorch_model-00003-of-00015.bin",
49
  "model.layers.12.mlp.down_proj.weight": "pytorch_model-00003-of-00015.bin",
@@ -53,7 +49,6 @@
53
  "model.layers.12.self_attn.k_proj.weight": "pytorch_model-00003-of-00015.bin",
54
  "model.layers.12.self_attn.o_proj.weight": "pytorch_model-00003-of-00015.bin",
55
  "model.layers.12.self_attn.q_proj.weight": "pytorch_model-00003-of-00015.bin",
56
- "model.layers.12.self_attn.rotary_emb.inv_freq": "pytorch_model-00003-of-00015.bin",
57
  "model.layers.12.self_attn.v_proj.weight": "pytorch_model-00003-of-00015.bin",
58
  "model.layers.13.input_layernorm.weight": "pytorch_model-00003-of-00015.bin",
59
  "model.layers.13.mlp.down_proj.weight": "pytorch_model-00003-of-00015.bin",
@@ -63,7 +58,6 @@
63
  "model.layers.13.self_attn.k_proj.weight": "pytorch_model-00003-of-00015.bin",
64
  "model.layers.13.self_attn.o_proj.weight": "pytorch_model-00003-of-00015.bin",
65
  "model.layers.13.self_attn.q_proj.weight": "pytorch_model-00003-of-00015.bin",
66
- "model.layers.13.self_attn.rotary_emb.inv_freq": "pytorch_model-00003-of-00015.bin",
67
  "model.layers.13.self_attn.v_proj.weight": "pytorch_model-00003-of-00015.bin",
68
  "model.layers.14.input_layernorm.weight": "pytorch_model-00003-of-00015.bin",
69
  "model.layers.14.mlp.down_proj.weight": "pytorch_model-00003-of-00015.bin",
@@ -73,7 +67,6 @@
73
  "model.layers.14.self_attn.k_proj.weight": "pytorch_model-00003-of-00015.bin",
74
  "model.layers.14.self_attn.o_proj.weight": "pytorch_model-00003-of-00015.bin",
75
  "model.layers.14.self_attn.q_proj.weight": "pytorch_model-00003-of-00015.bin",
76
- "model.layers.14.self_attn.rotary_emb.inv_freq": "pytorch_model-00003-of-00015.bin",
77
  "model.layers.14.self_attn.v_proj.weight": "pytorch_model-00003-of-00015.bin",
78
  "model.layers.15.input_layernorm.weight": "pytorch_model-00003-of-00015.bin",
79
  "model.layers.15.mlp.down_proj.weight": "pytorch_model-00003-of-00015.bin",
@@ -83,7 +76,6 @@
83
  "model.layers.15.self_attn.k_proj.weight": "pytorch_model-00003-of-00015.bin",
84
  "model.layers.15.self_attn.o_proj.weight": "pytorch_model-00003-of-00015.bin",
85
  "model.layers.15.self_attn.q_proj.weight": "pytorch_model-00003-of-00015.bin",
86
- "model.layers.15.self_attn.rotary_emb.inv_freq": "pytorch_model-00003-of-00015.bin",
87
  "model.layers.15.self_attn.v_proj.weight": "pytorch_model-00003-of-00015.bin",
88
  "model.layers.16.input_layernorm.weight": "pytorch_model-00003-of-00015.bin",
89
  "model.layers.16.mlp.down_proj.weight": "pytorch_model-00003-of-00015.bin",
@@ -93,7 +85,6 @@
93
  "model.layers.16.self_attn.k_proj.weight": "pytorch_model-00003-of-00015.bin",
94
  "model.layers.16.self_attn.o_proj.weight": "pytorch_model-00003-of-00015.bin",
95
  "model.layers.16.self_attn.q_proj.weight": "pytorch_model-00003-of-00015.bin",
96
- "model.layers.16.self_attn.rotary_emb.inv_freq": "pytorch_model-00003-of-00015.bin",
97
  "model.layers.16.self_attn.v_proj.weight": "pytorch_model-00003-of-00015.bin",
98
  "model.layers.17.input_layernorm.weight": "pytorch_model-00004-of-00015.bin",
99
  "model.layers.17.mlp.down_proj.weight": "pytorch_model-00004-of-00015.bin",
@@ -103,7 +94,6 @@
103
  "model.layers.17.self_attn.k_proj.weight": "pytorch_model-00004-of-00015.bin",
104
  "model.layers.17.self_attn.o_proj.weight": "pytorch_model-00004-of-00015.bin",
105
  "model.layers.17.self_attn.q_proj.weight": "pytorch_model-00004-of-00015.bin",
106
- "model.layers.17.self_attn.rotary_emb.inv_freq": "pytorch_model-00004-of-00015.bin",
107
  "model.layers.17.self_attn.v_proj.weight": "pytorch_model-00004-of-00015.bin",
108
  "model.layers.18.input_layernorm.weight": "pytorch_model-00004-of-00015.bin",
109
  "model.layers.18.mlp.down_proj.weight": "pytorch_model-00004-of-00015.bin",
@@ -113,7 +103,6 @@
113
  "model.layers.18.self_attn.k_proj.weight": "pytorch_model-00004-of-00015.bin",
114
  "model.layers.18.self_attn.o_proj.weight": "pytorch_model-00004-of-00015.bin",
115
  "model.layers.18.self_attn.q_proj.weight": "pytorch_model-00004-of-00015.bin",
116
- "model.layers.18.self_attn.rotary_emb.inv_freq": "pytorch_model-00004-of-00015.bin",
117
  "model.layers.18.self_attn.v_proj.weight": "pytorch_model-00004-of-00015.bin",
118
  "model.layers.19.input_layernorm.weight": "pytorch_model-00004-of-00015.bin",
119
  "model.layers.19.mlp.down_proj.weight": "pytorch_model-00004-of-00015.bin",
@@ -123,7 +112,6 @@
123
  "model.layers.19.self_attn.k_proj.weight": "pytorch_model-00004-of-00015.bin",
124
  "model.layers.19.self_attn.o_proj.weight": "pytorch_model-00004-of-00015.bin",
125
  "model.layers.19.self_attn.q_proj.weight": "pytorch_model-00004-of-00015.bin",
126
- "model.layers.19.self_attn.rotary_emb.inv_freq": "pytorch_model-00004-of-00015.bin",
127
  "model.layers.19.self_attn.v_proj.weight": "pytorch_model-00004-of-00015.bin",
128
  "model.layers.2.input_layernorm.weight": "pytorch_model-00001-of-00015.bin",
129
  "model.layers.2.mlp.down_proj.weight": "pytorch_model-00001-of-00015.bin",
@@ -133,7 +121,6 @@
133
  "model.layers.2.self_attn.k_proj.weight": "pytorch_model-00001-of-00015.bin",
134
  "model.layers.2.self_attn.o_proj.weight": "pytorch_model-00001-of-00015.bin",
135
  "model.layers.2.self_attn.q_proj.weight": "pytorch_model-00001-of-00015.bin",
136
- "model.layers.2.self_attn.rotary_emb.inv_freq": "pytorch_model-00001-of-00015.bin",
137
  "model.layers.2.self_attn.v_proj.weight": "pytorch_model-00001-of-00015.bin",
138
  "model.layers.20.input_layernorm.weight": "pytorch_model-00004-of-00015.bin",
139
  "model.layers.20.mlp.down_proj.weight": "pytorch_model-00004-of-00015.bin",
@@ -143,7 +130,6 @@
143
  "model.layers.20.self_attn.k_proj.weight": "pytorch_model-00004-of-00015.bin",
144
  "model.layers.20.self_attn.o_proj.weight": "pytorch_model-00004-of-00015.bin",
145
  "model.layers.20.self_attn.q_proj.weight": "pytorch_model-00004-of-00015.bin",
146
- "model.layers.20.self_attn.rotary_emb.inv_freq": "pytorch_model-00004-of-00015.bin",
147
  "model.layers.20.self_attn.v_proj.weight": "pytorch_model-00004-of-00015.bin",
148
  "model.layers.21.input_layernorm.weight": "pytorch_model-00004-of-00015.bin",
149
  "model.layers.21.mlp.down_proj.weight": "pytorch_model-00004-of-00015.bin",
@@ -153,7 +139,6 @@
153
  "model.layers.21.self_attn.k_proj.weight": "pytorch_model-00004-of-00015.bin",
154
  "model.layers.21.self_attn.o_proj.weight": "pytorch_model-00004-of-00015.bin",
155
  "model.layers.21.self_attn.q_proj.weight": "pytorch_model-00004-of-00015.bin",
156
- "model.layers.21.self_attn.rotary_emb.inv_freq": "pytorch_model-00004-of-00015.bin",
157
  "model.layers.21.self_attn.v_proj.weight": "pytorch_model-00004-of-00015.bin",
158
  "model.layers.22.input_layernorm.weight": "pytorch_model-00005-of-00015.bin",
159
  "model.layers.22.mlp.down_proj.weight": "pytorch_model-00005-of-00015.bin",
@@ -163,7 +148,6 @@
163
  "model.layers.22.self_attn.k_proj.weight": "pytorch_model-00004-of-00015.bin",
164
  "model.layers.22.self_attn.o_proj.weight": "pytorch_model-00004-of-00015.bin",
165
  "model.layers.22.self_attn.q_proj.weight": "pytorch_model-00004-of-00015.bin",
166
- "model.layers.22.self_attn.rotary_emb.inv_freq": "pytorch_model-00004-of-00015.bin",
167
  "model.layers.22.self_attn.v_proj.weight": "pytorch_model-00004-of-00015.bin",
168
  "model.layers.23.input_layernorm.weight": "pytorch_model-00005-of-00015.bin",
169
  "model.layers.23.mlp.down_proj.weight": "pytorch_model-00005-of-00015.bin",
@@ -173,7 +157,6 @@
173
  "model.layers.23.self_attn.k_proj.weight": "pytorch_model-00005-of-00015.bin",
174
  "model.layers.23.self_attn.o_proj.weight": "pytorch_model-00005-of-00015.bin",
175
  "model.layers.23.self_attn.q_proj.weight": "pytorch_model-00005-of-00015.bin",
176
- "model.layers.23.self_attn.rotary_emb.inv_freq": "pytorch_model-00005-of-00015.bin",
177
  "model.layers.23.self_attn.v_proj.weight": "pytorch_model-00005-of-00015.bin",
178
  "model.layers.24.input_layernorm.weight": "pytorch_model-00005-of-00015.bin",
179
  "model.layers.24.mlp.down_proj.weight": "pytorch_model-00005-of-00015.bin",
@@ -183,7 +166,6 @@
183
  "model.layers.24.self_attn.k_proj.weight": "pytorch_model-00005-of-00015.bin",
184
  "model.layers.24.self_attn.o_proj.weight": "pytorch_model-00005-of-00015.bin",
185
  "model.layers.24.self_attn.q_proj.weight": "pytorch_model-00005-of-00015.bin",
186
- "model.layers.24.self_attn.rotary_emb.inv_freq": "pytorch_model-00005-of-00015.bin",
187
  "model.layers.24.self_attn.v_proj.weight": "pytorch_model-00005-of-00015.bin",
188
  "model.layers.25.input_layernorm.weight": "pytorch_model-00005-of-00015.bin",
189
  "model.layers.25.mlp.down_proj.weight": "pytorch_model-00005-of-00015.bin",
@@ -193,7 +175,6 @@
193
  "model.layers.25.self_attn.k_proj.weight": "pytorch_model-00005-of-00015.bin",
194
  "model.layers.25.self_attn.o_proj.weight": "pytorch_model-00005-of-00015.bin",
195
  "model.layers.25.self_attn.q_proj.weight": "pytorch_model-00005-of-00015.bin",
196
- "model.layers.25.self_attn.rotary_emb.inv_freq": "pytorch_model-00005-of-00015.bin",
197
  "model.layers.25.self_attn.v_proj.weight": "pytorch_model-00005-of-00015.bin",
198
  "model.layers.26.input_layernorm.weight": "pytorch_model-00005-of-00015.bin",
199
  "model.layers.26.mlp.down_proj.weight": "pytorch_model-00005-of-00015.bin",
@@ -203,7 +184,6 @@
203
  "model.layers.26.self_attn.k_proj.weight": "pytorch_model-00005-of-00015.bin",
204
  "model.layers.26.self_attn.o_proj.weight": "pytorch_model-00005-of-00015.bin",
205
  "model.layers.26.self_attn.q_proj.weight": "pytorch_model-00005-of-00015.bin",
206
- "model.layers.26.self_attn.rotary_emb.inv_freq": "pytorch_model-00005-of-00015.bin",
207
  "model.layers.26.self_attn.v_proj.weight": "pytorch_model-00005-of-00015.bin",
208
  "model.layers.27.input_layernorm.weight": "pytorch_model-00005-of-00015.bin",
209
  "model.layers.27.mlp.down_proj.weight": "pytorch_model-00005-of-00015.bin",
@@ -213,7 +193,6 @@
213
  "model.layers.27.self_attn.k_proj.weight": "pytorch_model-00005-of-00015.bin",
214
  "model.layers.27.self_attn.o_proj.weight": "pytorch_model-00005-of-00015.bin",
215
  "model.layers.27.self_attn.q_proj.weight": "pytorch_model-00005-of-00015.bin",
216
- "model.layers.27.self_attn.rotary_emb.inv_freq": "pytorch_model-00005-of-00015.bin",
217
  "model.layers.27.self_attn.v_proj.weight": "pytorch_model-00005-of-00015.bin",
218
  "model.layers.28.input_layernorm.weight": "pytorch_model-00006-of-00015.bin",
219
  "model.layers.28.mlp.down_proj.weight": "pytorch_model-00006-of-00015.bin",
@@ -223,7 +202,6 @@
223
  "model.layers.28.self_attn.k_proj.weight": "pytorch_model-00005-of-00015.bin",
224
  "model.layers.28.self_attn.o_proj.weight": "pytorch_model-00005-of-00015.bin",
225
  "model.layers.28.self_attn.q_proj.weight": "pytorch_model-00005-of-00015.bin",
226
- "model.layers.28.self_attn.rotary_emb.inv_freq": "pytorch_model-00005-of-00015.bin",
227
  "model.layers.28.self_attn.v_proj.weight": "pytorch_model-00005-of-00015.bin",
228
  "model.layers.29.input_layernorm.weight": "pytorch_model-00006-of-00015.bin",
229
  "model.layers.29.mlp.down_proj.weight": "pytorch_model-00006-of-00015.bin",
@@ -233,7 +211,6 @@
233
  "model.layers.29.self_attn.k_proj.weight": "pytorch_model-00006-of-00015.bin",
234
  "model.layers.29.self_attn.o_proj.weight": "pytorch_model-00006-of-00015.bin",
235
  "model.layers.29.self_attn.q_proj.weight": "pytorch_model-00006-of-00015.bin",
236
- "model.layers.29.self_attn.rotary_emb.inv_freq": "pytorch_model-00006-of-00015.bin",
237
  "model.layers.29.self_attn.v_proj.weight": "pytorch_model-00006-of-00015.bin",
238
  "model.layers.3.input_layernorm.weight": "pytorch_model-00001-of-00015.bin",
239
  "model.layers.3.mlp.down_proj.weight": "pytorch_model-00001-of-00015.bin",
@@ -243,7 +220,6 @@
243
  "model.layers.3.self_attn.k_proj.weight": "pytorch_model-00001-of-00015.bin",
244
  "model.layers.3.self_attn.o_proj.weight": "pytorch_model-00001-of-00015.bin",
245
  "model.layers.3.self_attn.q_proj.weight": "pytorch_model-00001-of-00015.bin",
246
- "model.layers.3.self_attn.rotary_emb.inv_freq": "pytorch_model-00001-of-00015.bin",
247
  "model.layers.3.self_attn.v_proj.weight": "pytorch_model-00001-of-00015.bin",
248
  "model.layers.30.input_layernorm.weight": "pytorch_model-00006-of-00015.bin",
249
  "model.layers.30.mlp.down_proj.weight": "pytorch_model-00006-of-00015.bin",
@@ -253,7 +229,6 @@
253
  "model.layers.30.self_attn.k_proj.weight": "pytorch_model-00006-of-00015.bin",
254
  "model.layers.30.self_attn.o_proj.weight": "pytorch_model-00006-of-00015.bin",
255
  "model.layers.30.self_attn.q_proj.weight": "pytorch_model-00006-of-00015.bin",
256
- "model.layers.30.self_attn.rotary_emb.inv_freq": "pytorch_model-00006-of-00015.bin",
257
  "model.layers.30.self_attn.v_proj.weight": "pytorch_model-00006-of-00015.bin",
258
  "model.layers.31.input_layernorm.weight": "pytorch_model-00006-of-00015.bin",
259
  "model.layers.31.mlp.down_proj.weight": "pytorch_model-00006-of-00015.bin",
@@ -263,7 +238,6 @@
263
  "model.layers.31.self_attn.k_proj.weight": "pytorch_model-00006-of-00015.bin",
264
  "model.layers.31.self_attn.o_proj.weight": "pytorch_model-00006-of-00015.bin",
265
  "model.layers.31.self_attn.q_proj.weight": "pytorch_model-00006-of-00015.bin",
266
- "model.layers.31.self_attn.rotary_emb.inv_freq": "pytorch_model-00006-of-00015.bin",
267
  "model.layers.31.self_attn.v_proj.weight": "pytorch_model-00006-of-00015.bin",
268
  "model.layers.32.input_layernorm.weight": "pytorch_model-00006-of-00015.bin",
269
  "model.layers.32.mlp.down_proj.weight": "pytorch_model-00006-of-00015.bin",
@@ -273,7 +247,6 @@
273
  "model.layers.32.self_attn.k_proj.weight": "pytorch_model-00006-of-00015.bin",
274
  "model.layers.32.self_attn.o_proj.weight": "pytorch_model-00006-of-00015.bin",
275
  "model.layers.32.self_attn.q_proj.weight": "pytorch_model-00006-of-00015.bin",
276
- "model.layers.32.self_attn.rotary_emb.inv_freq": "pytorch_model-00006-of-00015.bin",
277
  "model.layers.32.self_attn.v_proj.weight": "pytorch_model-00006-of-00015.bin",
278
  "model.layers.33.input_layernorm.weight": "pytorch_model-00006-of-00015.bin",
279
  "model.layers.33.mlp.down_proj.weight": "pytorch_model-00006-of-00015.bin",
@@ -283,7 +256,6 @@
283
  "model.layers.33.self_attn.k_proj.weight": "pytorch_model-00006-of-00015.bin",
284
  "model.layers.33.self_attn.o_proj.weight": "pytorch_model-00006-of-00015.bin",
285
  "model.layers.33.self_attn.q_proj.weight": "pytorch_model-00006-of-00015.bin",
286
- "model.layers.33.self_attn.rotary_emb.inv_freq": "pytorch_model-00006-of-00015.bin",
287
  "model.layers.33.self_attn.v_proj.weight": "pytorch_model-00006-of-00015.bin",
288
  "model.layers.34.input_layernorm.weight": "pytorch_model-00007-of-00015.bin",
289
  "model.layers.34.mlp.down_proj.weight": "pytorch_model-00007-of-00015.bin",
@@ -293,7 +265,6 @@
293
  "model.layers.34.self_attn.k_proj.weight": "pytorch_model-00006-of-00015.bin",
294
  "model.layers.34.self_attn.o_proj.weight": "pytorch_model-00006-of-00015.bin",
295
  "model.layers.34.self_attn.q_proj.weight": "pytorch_model-00006-of-00015.bin",
296
- "model.layers.34.self_attn.rotary_emb.inv_freq": "pytorch_model-00006-of-00015.bin",
297
  "model.layers.34.self_attn.v_proj.weight": "pytorch_model-00006-of-00015.bin",
298
  "model.layers.35.input_layernorm.weight": "pytorch_model-00007-of-00015.bin",
299
  "model.layers.35.mlp.down_proj.weight": "pytorch_model-00007-of-00015.bin",
@@ -303,7 +274,6 @@
303
  "model.layers.35.self_attn.k_proj.weight": "pytorch_model-00007-of-00015.bin",
304
  "model.layers.35.self_attn.o_proj.weight": "pytorch_model-00007-of-00015.bin",
305
  "model.layers.35.self_attn.q_proj.weight": "pytorch_model-00007-of-00015.bin",
306
- "model.layers.35.self_attn.rotary_emb.inv_freq": "pytorch_model-00007-of-00015.bin",
307
  "model.layers.35.self_attn.v_proj.weight": "pytorch_model-00007-of-00015.bin",
308
  "model.layers.36.input_layernorm.weight": "pytorch_model-00007-of-00015.bin",
309
  "model.layers.36.mlp.down_proj.weight": "pytorch_model-00007-of-00015.bin",
@@ -313,7 +283,6 @@
313
  "model.layers.36.self_attn.k_proj.weight": "pytorch_model-00007-of-00015.bin",
314
  "model.layers.36.self_attn.o_proj.weight": "pytorch_model-00007-of-00015.bin",
315
  "model.layers.36.self_attn.q_proj.weight": "pytorch_model-00007-of-00015.bin",
316
- "model.layers.36.self_attn.rotary_emb.inv_freq": "pytorch_model-00007-of-00015.bin",
317
  "model.layers.36.self_attn.v_proj.weight": "pytorch_model-00007-of-00015.bin",
318
  "model.layers.37.input_layernorm.weight": "pytorch_model-00007-of-00015.bin",
319
  "model.layers.37.mlp.down_proj.weight": "pytorch_model-00007-of-00015.bin",
@@ -323,7 +292,6 @@
323
  "model.layers.37.self_attn.k_proj.weight": "pytorch_model-00007-of-00015.bin",
324
  "model.layers.37.self_attn.o_proj.weight": "pytorch_model-00007-of-00015.bin",
325
  "model.layers.37.self_attn.q_proj.weight": "pytorch_model-00007-of-00015.bin",
326
- "model.layers.37.self_attn.rotary_emb.inv_freq": "pytorch_model-00007-of-00015.bin",
327
  "model.layers.37.self_attn.v_proj.weight": "pytorch_model-00007-of-00015.bin",
328
  "model.layers.38.input_layernorm.weight": "pytorch_model-00007-of-00015.bin",
329
  "model.layers.38.mlp.down_proj.weight": "pytorch_model-00007-of-00015.bin",
@@ -333,7 +301,6 @@
333
  "model.layers.38.self_attn.k_proj.weight": "pytorch_model-00007-of-00015.bin",
334
  "model.layers.38.self_attn.o_proj.weight": "pytorch_model-00007-of-00015.bin",
335
  "model.layers.38.self_attn.q_proj.weight": "pytorch_model-00007-of-00015.bin",
336
- "model.layers.38.self_attn.rotary_emb.inv_freq": "pytorch_model-00007-of-00015.bin",
337
  "model.layers.38.self_attn.v_proj.weight": "pytorch_model-00007-of-00015.bin",
338
  "model.layers.39.input_layernorm.weight": "pytorch_model-00007-of-00015.bin",
339
  "model.layers.39.mlp.down_proj.weight": "pytorch_model-00007-of-00015.bin",
@@ -343,7 +310,6 @@
343
  "model.layers.39.self_attn.k_proj.weight": "pytorch_model-00007-of-00015.bin",
344
  "model.layers.39.self_attn.o_proj.weight": "pytorch_model-00007-of-00015.bin",
345
  "model.layers.39.self_attn.q_proj.weight": "pytorch_model-00007-of-00015.bin",
346
- "model.layers.39.self_attn.rotary_emb.inv_freq": "pytorch_model-00007-of-00015.bin",
347
  "model.layers.39.self_attn.v_proj.weight": "pytorch_model-00007-of-00015.bin",
348
  "model.layers.4.input_layernorm.weight": "pytorch_model-00001-of-00015.bin",
349
  "model.layers.4.mlp.down_proj.weight": "pytorch_model-00001-of-00015.bin",
@@ -353,7 +319,6 @@
353
  "model.layers.4.self_attn.k_proj.weight": "pytorch_model-00001-of-00015.bin",
354
  "model.layers.4.self_attn.o_proj.weight": "pytorch_model-00001-of-00015.bin",
355
  "model.layers.4.self_attn.q_proj.weight": "pytorch_model-00001-of-00015.bin",
356
- "model.layers.4.self_attn.rotary_emb.inv_freq": "pytorch_model-00001-of-00015.bin",
357
  "model.layers.4.self_attn.v_proj.weight": "pytorch_model-00001-of-00015.bin",
358
  "model.layers.40.input_layernorm.weight": "pytorch_model-00008-of-00015.bin",
359
  "model.layers.40.mlp.down_proj.weight": "pytorch_model-00008-of-00015.bin",
@@ -363,7 +328,6 @@
363
  "model.layers.40.self_attn.k_proj.weight": "pytorch_model-00008-of-00015.bin",
364
  "model.layers.40.self_attn.o_proj.weight": "pytorch_model-00008-of-00015.bin",
365
  "model.layers.40.self_attn.q_proj.weight": "pytorch_model-00008-of-00015.bin",
366
- "model.layers.40.self_attn.rotary_emb.inv_freq": "pytorch_model-00008-of-00015.bin",
367
  "model.layers.40.self_attn.v_proj.weight": "pytorch_model-00008-of-00015.bin",
368
  "model.layers.41.input_layernorm.weight": "pytorch_model-00008-of-00015.bin",
369
  "model.layers.41.mlp.down_proj.weight": "pytorch_model-00008-of-00015.bin",
@@ -373,7 +337,6 @@
373
  "model.layers.41.self_attn.k_proj.weight": "pytorch_model-00008-of-00015.bin",
374
  "model.layers.41.self_attn.o_proj.weight": "pytorch_model-00008-of-00015.bin",
375
  "model.layers.41.self_attn.q_proj.weight": "pytorch_model-00008-of-00015.bin",
376
- "model.layers.41.self_attn.rotary_emb.inv_freq": "pytorch_model-00008-of-00015.bin",
377
  "model.layers.41.self_attn.v_proj.weight": "pytorch_model-00008-of-00015.bin",
378
  "model.layers.42.input_layernorm.weight": "pytorch_model-00008-of-00015.bin",
379
  "model.layers.42.mlp.down_proj.weight": "pytorch_model-00008-of-00015.bin",
@@ -383,7 +346,6 @@
383
  "model.layers.42.self_attn.k_proj.weight": "pytorch_model-00008-of-00015.bin",
384
  "model.layers.42.self_attn.o_proj.weight": "pytorch_model-00008-of-00015.bin",
385
  "model.layers.42.self_attn.q_proj.weight": "pytorch_model-00008-of-00015.bin",
386
- "model.layers.42.self_attn.rotary_emb.inv_freq": "pytorch_model-00008-of-00015.bin",
387
  "model.layers.42.self_attn.v_proj.weight": "pytorch_model-00008-of-00015.bin",
388
  "model.layers.43.input_layernorm.weight": "pytorch_model-00008-of-00015.bin",
389
  "model.layers.43.mlp.down_proj.weight": "pytorch_model-00008-of-00015.bin",
@@ -393,7 +355,6 @@
393
  "model.layers.43.self_attn.k_proj.weight": "pytorch_model-00008-of-00015.bin",
394
  "model.layers.43.self_attn.o_proj.weight": "pytorch_model-00008-of-00015.bin",
395
  "model.layers.43.self_attn.q_proj.weight": "pytorch_model-00008-of-00015.bin",
396
- "model.layers.43.self_attn.rotary_emb.inv_freq": "pytorch_model-00008-of-00015.bin",
397
  "model.layers.43.self_attn.v_proj.weight": "pytorch_model-00008-of-00015.bin",
398
  "model.layers.44.input_layernorm.weight": "pytorch_model-00008-of-00015.bin",
399
  "model.layers.44.mlp.down_proj.weight": "pytorch_model-00008-of-00015.bin",
@@ -403,7 +364,6 @@
403
  "model.layers.44.self_attn.k_proj.weight": "pytorch_model-00008-of-00015.bin",
404
  "model.layers.44.self_attn.o_proj.weight": "pytorch_model-00008-of-00015.bin",
405
  "model.layers.44.self_attn.q_proj.weight": "pytorch_model-00008-of-00015.bin",
406
- "model.layers.44.self_attn.rotary_emb.inv_freq": "pytorch_model-00008-of-00015.bin",
407
  "model.layers.44.self_attn.v_proj.weight": "pytorch_model-00008-of-00015.bin",
408
  "model.layers.45.input_layernorm.weight": "pytorch_model-00009-of-00015.bin",
409
  "model.layers.45.mlp.down_proj.weight": "pytorch_model-00009-of-00015.bin",
@@ -413,7 +373,6 @@
413
  "model.layers.45.self_attn.k_proj.weight": "pytorch_model-00008-of-00015.bin",
414
  "model.layers.45.self_attn.o_proj.weight": "pytorch_model-00008-of-00015.bin",
415
  "model.layers.45.self_attn.q_proj.weight": "pytorch_model-00008-of-00015.bin",
416
- "model.layers.45.self_attn.rotary_emb.inv_freq": "pytorch_model-00008-of-00015.bin",
417
  "model.layers.45.self_attn.v_proj.weight": "pytorch_model-00008-of-00015.bin",
418
  "model.layers.46.input_layernorm.weight": "pytorch_model-00009-of-00015.bin",
419
  "model.layers.46.mlp.down_proj.weight": "pytorch_model-00009-of-00015.bin",
@@ -423,7 +382,6 @@
423
  "model.layers.46.self_attn.k_proj.weight": "pytorch_model-00009-of-00015.bin",
424
  "model.layers.46.self_attn.o_proj.weight": "pytorch_model-00009-of-00015.bin",
425
  "model.layers.46.self_attn.q_proj.weight": "pytorch_model-00009-of-00015.bin",
426
- "model.layers.46.self_attn.rotary_emb.inv_freq": "pytorch_model-00009-of-00015.bin",
427
  "model.layers.46.self_attn.v_proj.weight": "pytorch_model-00009-of-00015.bin",
428
  "model.layers.47.input_layernorm.weight": "pytorch_model-00009-of-00015.bin",
429
  "model.layers.47.mlp.down_proj.weight": "pytorch_model-00009-of-00015.bin",
@@ -433,7 +391,6 @@
433
  "model.layers.47.self_attn.k_proj.weight": "pytorch_model-00009-of-00015.bin",
434
  "model.layers.47.self_attn.o_proj.weight": "pytorch_model-00009-of-00015.bin",
435
  "model.layers.47.self_attn.q_proj.weight": "pytorch_model-00009-of-00015.bin",
436
- "model.layers.47.self_attn.rotary_emb.inv_freq": "pytorch_model-00009-of-00015.bin",
437
  "model.layers.47.self_attn.v_proj.weight": "pytorch_model-00009-of-00015.bin",
438
  "model.layers.48.input_layernorm.weight": "pytorch_model-00009-of-00015.bin",
439
  "model.layers.48.mlp.down_proj.weight": "pytorch_model-00009-of-00015.bin",
@@ -443,7 +400,6 @@
443
  "model.layers.48.self_attn.k_proj.weight": "pytorch_model-00009-of-00015.bin",
444
  "model.layers.48.self_attn.o_proj.weight": "pytorch_model-00009-of-00015.bin",
445
  "model.layers.48.self_attn.q_proj.weight": "pytorch_model-00009-of-00015.bin",
446
- "model.layers.48.self_attn.rotary_emb.inv_freq": "pytorch_model-00009-of-00015.bin",
447
  "model.layers.48.self_attn.v_proj.weight": "pytorch_model-00009-of-00015.bin",
448
  "model.layers.49.input_layernorm.weight": "pytorch_model-00009-of-00015.bin",
449
  "model.layers.49.mlp.down_proj.weight": "pytorch_model-00009-of-00015.bin",
@@ -453,7 +409,6 @@
453
  "model.layers.49.self_attn.k_proj.weight": "pytorch_model-00009-of-00015.bin",
454
  "model.layers.49.self_attn.o_proj.weight": "pytorch_model-00009-of-00015.bin",
455
  "model.layers.49.self_attn.q_proj.weight": "pytorch_model-00009-of-00015.bin",
456
- "model.layers.49.self_attn.rotary_emb.inv_freq": "pytorch_model-00009-of-00015.bin",
457
  "model.layers.49.self_attn.v_proj.weight": "pytorch_model-00009-of-00015.bin",
458
  "model.layers.5.input_layernorm.weight": "pytorch_model-00002-of-00015.bin",
459
  "model.layers.5.mlp.down_proj.weight": "pytorch_model-00002-of-00015.bin",
@@ -463,7 +418,6 @@
463
  "model.layers.5.self_attn.k_proj.weight": "pytorch_model-00001-of-00015.bin",
464
  "model.layers.5.self_attn.o_proj.weight": "pytorch_model-00001-of-00015.bin",
465
  "model.layers.5.self_attn.q_proj.weight": "pytorch_model-00001-of-00015.bin",
466
- "model.layers.5.self_attn.rotary_emb.inv_freq": "pytorch_model-00001-of-00015.bin",
467
  "model.layers.5.self_attn.v_proj.weight": "pytorch_model-00001-of-00015.bin",
468
  "model.layers.50.input_layernorm.weight": "pytorch_model-00009-of-00015.bin",
469
  "model.layers.50.mlp.down_proj.weight": "pytorch_model-00009-of-00015.bin",
@@ -473,7 +427,6 @@
473
  "model.layers.50.self_attn.k_proj.weight": "pytorch_model-00009-of-00015.bin",
474
  "model.layers.50.self_attn.o_proj.weight": "pytorch_model-00009-of-00015.bin",
475
  "model.layers.50.self_attn.q_proj.weight": "pytorch_model-00009-of-00015.bin",
476
- "model.layers.50.self_attn.rotary_emb.inv_freq": "pytorch_model-00009-of-00015.bin",
477
  "model.layers.50.self_attn.v_proj.weight": "pytorch_model-00009-of-00015.bin",
478
  "model.layers.51.input_layernorm.weight": "pytorch_model-00010-of-00015.bin",
479
  "model.layers.51.mlp.down_proj.weight": "pytorch_model-00010-of-00015.bin",
@@ -483,7 +436,6 @@
483
  "model.layers.51.self_attn.k_proj.weight": "pytorch_model-00009-of-00015.bin",
484
  "model.layers.51.self_attn.o_proj.weight": "pytorch_model-00009-of-00015.bin",
485
  "model.layers.51.self_attn.q_proj.weight": "pytorch_model-00009-of-00015.bin",
486
- "model.layers.51.self_attn.rotary_emb.inv_freq": "pytorch_model-00009-of-00015.bin",
487
  "model.layers.51.self_attn.v_proj.weight": "pytorch_model-00009-of-00015.bin",
488
  "model.layers.52.input_layernorm.weight": "pytorch_model-00010-of-00015.bin",
489
  "model.layers.52.mlp.down_proj.weight": "pytorch_model-00010-of-00015.bin",
@@ -493,7 +445,6 @@
493
  "model.layers.52.self_attn.k_proj.weight": "pytorch_model-00010-of-00015.bin",
494
  "model.layers.52.self_attn.o_proj.weight": "pytorch_model-00010-of-00015.bin",
495
  "model.layers.52.self_attn.q_proj.weight": "pytorch_model-00010-of-00015.bin",
496
- "model.layers.52.self_attn.rotary_emb.inv_freq": "pytorch_model-00010-of-00015.bin",
497
  "model.layers.52.self_attn.v_proj.weight": "pytorch_model-00010-of-00015.bin",
498
  "model.layers.53.input_layernorm.weight": "pytorch_model-00010-of-00015.bin",
499
  "model.layers.53.mlp.down_proj.weight": "pytorch_model-00010-of-00015.bin",
@@ -503,7 +454,6 @@
503
  "model.layers.53.self_attn.k_proj.weight": "pytorch_model-00010-of-00015.bin",
504
  "model.layers.53.self_attn.o_proj.weight": "pytorch_model-00010-of-00015.bin",
505
  "model.layers.53.self_attn.q_proj.weight": "pytorch_model-00010-of-00015.bin",
506
- "model.layers.53.self_attn.rotary_emb.inv_freq": "pytorch_model-00010-of-00015.bin",
507
  "model.layers.53.self_attn.v_proj.weight": "pytorch_model-00010-of-00015.bin",
508
  "model.layers.54.input_layernorm.weight": "pytorch_model-00010-of-00015.bin",
509
  "model.layers.54.mlp.down_proj.weight": "pytorch_model-00010-of-00015.bin",
@@ -513,7 +463,6 @@
513
  "model.layers.54.self_attn.k_proj.weight": "pytorch_model-00010-of-00015.bin",
514
  "model.layers.54.self_attn.o_proj.weight": "pytorch_model-00010-of-00015.bin",
515
  "model.layers.54.self_attn.q_proj.weight": "pytorch_model-00010-of-00015.bin",
516
- "model.layers.54.self_attn.rotary_emb.inv_freq": "pytorch_model-00010-of-00015.bin",
517
  "model.layers.54.self_attn.v_proj.weight": "pytorch_model-00010-of-00015.bin",
518
  "model.layers.55.input_layernorm.weight": "pytorch_model-00010-of-00015.bin",
519
  "model.layers.55.mlp.down_proj.weight": "pytorch_model-00010-of-00015.bin",
@@ -523,7 +472,6 @@
523
  "model.layers.55.self_attn.k_proj.weight": "pytorch_model-00010-of-00015.bin",
524
  "model.layers.55.self_attn.o_proj.weight": "pytorch_model-00010-of-00015.bin",
525
  "model.layers.55.self_attn.q_proj.weight": "pytorch_model-00010-of-00015.bin",
526
- "model.layers.55.self_attn.rotary_emb.inv_freq": "pytorch_model-00010-of-00015.bin",
527
  "model.layers.55.self_attn.v_proj.weight": "pytorch_model-00010-of-00015.bin",
528
  "model.layers.56.input_layernorm.weight": "pytorch_model-00010-of-00015.bin",
529
  "model.layers.56.mlp.down_proj.weight": "pytorch_model-00010-of-00015.bin",
@@ -533,7 +481,6 @@
533
  "model.layers.56.self_attn.k_proj.weight": "pytorch_model-00010-of-00015.bin",
534
  "model.layers.56.self_attn.o_proj.weight": "pytorch_model-00010-of-00015.bin",
535
  "model.layers.56.self_attn.q_proj.weight": "pytorch_model-00010-of-00015.bin",
536
- "model.layers.56.self_attn.rotary_emb.inv_freq": "pytorch_model-00010-of-00015.bin",
537
  "model.layers.56.self_attn.v_proj.weight": "pytorch_model-00010-of-00015.bin",
538
  "model.layers.57.input_layernorm.weight": "pytorch_model-00011-of-00015.bin",
539
  "model.layers.57.mlp.down_proj.weight": "pytorch_model-00011-of-00015.bin",
@@ -543,7 +490,6 @@
543
  "model.layers.57.self_attn.k_proj.weight": "pytorch_model-00010-of-00015.bin",
544
  "model.layers.57.self_attn.o_proj.weight": "pytorch_model-00010-of-00015.bin",
545
  "model.layers.57.self_attn.q_proj.weight": "pytorch_model-00010-of-00015.bin",
546
- "model.layers.57.self_attn.rotary_emb.inv_freq": "pytorch_model-00010-of-00015.bin",
547
  "model.layers.57.self_attn.v_proj.weight": "pytorch_model-00010-of-00015.bin",
548
  "model.layers.58.input_layernorm.weight": "pytorch_model-00011-of-00015.bin",
549
  "model.layers.58.mlp.down_proj.weight": "pytorch_model-00011-of-00015.bin",
@@ -553,7 +499,6 @@
553
  "model.layers.58.self_attn.k_proj.weight": "pytorch_model-00011-of-00015.bin",
554
  "model.layers.58.self_attn.o_proj.weight": "pytorch_model-00011-of-00015.bin",
555
  "model.layers.58.self_attn.q_proj.weight": "pytorch_model-00011-of-00015.bin",
556
- "model.layers.58.self_attn.rotary_emb.inv_freq": "pytorch_model-00011-of-00015.bin",
557
  "model.layers.58.self_attn.v_proj.weight": "pytorch_model-00011-of-00015.bin",
558
  "model.layers.59.input_layernorm.weight": "pytorch_model-00011-of-00015.bin",
559
  "model.layers.59.mlp.down_proj.weight": "pytorch_model-00011-of-00015.bin",
@@ -563,7 +508,6 @@
563
  "model.layers.59.self_attn.k_proj.weight": "pytorch_model-00011-of-00015.bin",
564
  "model.layers.59.self_attn.o_proj.weight": "pytorch_model-00011-of-00015.bin",
565
  "model.layers.59.self_attn.q_proj.weight": "pytorch_model-00011-of-00015.bin",
566
- "model.layers.59.self_attn.rotary_emb.inv_freq": "pytorch_model-00011-of-00015.bin",
567
  "model.layers.59.self_attn.v_proj.weight": "pytorch_model-00011-of-00015.bin",
568
  "model.layers.6.input_layernorm.weight": "pytorch_model-00002-of-00015.bin",
569
  "model.layers.6.mlp.down_proj.weight": "pytorch_model-00002-of-00015.bin",
@@ -573,7 +517,6 @@
573
  "model.layers.6.self_attn.k_proj.weight": "pytorch_model-00002-of-00015.bin",
574
  "model.layers.6.self_attn.o_proj.weight": "pytorch_model-00002-of-00015.bin",
575
  "model.layers.6.self_attn.q_proj.weight": "pytorch_model-00002-of-00015.bin",
576
- "model.layers.6.self_attn.rotary_emb.inv_freq": "pytorch_model-00002-of-00015.bin",
577
  "model.layers.6.self_attn.v_proj.weight": "pytorch_model-00002-of-00015.bin",
578
  "model.layers.60.input_layernorm.weight": "pytorch_model-00011-of-00015.bin",
579
  "model.layers.60.mlp.down_proj.weight": "pytorch_model-00011-of-00015.bin",
@@ -583,7 +526,6 @@
583
  "model.layers.60.self_attn.k_proj.weight": "pytorch_model-00011-of-00015.bin",
584
  "model.layers.60.self_attn.o_proj.weight": "pytorch_model-00011-of-00015.bin",
585
  "model.layers.60.self_attn.q_proj.weight": "pytorch_model-00011-of-00015.bin",
586
- "model.layers.60.self_attn.rotary_emb.inv_freq": "pytorch_model-00011-of-00015.bin",
587
  "model.layers.60.self_attn.v_proj.weight": "pytorch_model-00011-of-00015.bin",
588
  "model.layers.61.input_layernorm.weight": "pytorch_model-00011-of-00015.bin",
589
  "model.layers.61.mlp.down_proj.weight": "pytorch_model-00011-of-00015.bin",
@@ -593,7 +535,6 @@
593
  "model.layers.61.self_attn.k_proj.weight": "pytorch_model-00011-of-00015.bin",
594
  "model.layers.61.self_attn.o_proj.weight": "pytorch_model-00011-of-00015.bin",
595
  "model.layers.61.self_attn.q_proj.weight": "pytorch_model-00011-of-00015.bin",
596
- "model.layers.61.self_attn.rotary_emb.inv_freq": "pytorch_model-00011-of-00015.bin",
597
  "model.layers.61.self_attn.v_proj.weight": "pytorch_model-00011-of-00015.bin",
598
  "model.layers.62.input_layernorm.weight": "pytorch_model-00011-of-00015.bin",
599
  "model.layers.62.mlp.down_proj.weight": "pytorch_model-00011-of-00015.bin",
@@ -603,7 +544,6 @@
603
  "model.layers.62.self_attn.k_proj.weight": "pytorch_model-00011-of-00015.bin",
604
  "model.layers.62.self_attn.o_proj.weight": "pytorch_model-00011-of-00015.bin",
605
  "model.layers.62.self_attn.q_proj.weight": "pytorch_model-00011-of-00015.bin",
606
- "model.layers.62.self_attn.rotary_emb.inv_freq": "pytorch_model-00011-of-00015.bin",
607
  "model.layers.62.self_attn.v_proj.weight": "pytorch_model-00011-of-00015.bin",
608
  "model.layers.63.input_layernorm.weight": "pytorch_model-00012-of-00015.bin",
609
  "model.layers.63.mlp.down_proj.weight": "pytorch_model-00012-of-00015.bin",
@@ -613,7 +553,6 @@
613
  "model.layers.63.self_attn.k_proj.weight": "pytorch_model-00012-of-00015.bin",
614
  "model.layers.63.self_attn.o_proj.weight": "pytorch_model-00012-of-00015.bin",
615
  "model.layers.63.self_attn.q_proj.weight": "pytorch_model-00012-of-00015.bin",
616
- "model.layers.63.self_attn.rotary_emb.inv_freq": "pytorch_model-00012-of-00015.bin",
617
  "model.layers.63.self_attn.v_proj.weight": "pytorch_model-00012-of-00015.bin",
618
  "model.layers.64.input_layernorm.weight": "pytorch_model-00012-of-00015.bin",
619
  "model.layers.64.mlp.down_proj.weight": "pytorch_model-00012-of-00015.bin",
@@ -623,7 +562,6 @@
623
  "model.layers.64.self_attn.k_proj.weight": "pytorch_model-00012-of-00015.bin",
624
  "model.layers.64.self_attn.o_proj.weight": "pytorch_model-00012-of-00015.bin",
625
  "model.layers.64.self_attn.q_proj.weight": "pytorch_model-00012-of-00015.bin",
626
- "model.layers.64.self_attn.rotary_emb.inv_freq": "pytorch_model-00012-of-00015.bin",
627
  "model.layers.64.self_attn.v_proj.weight": "pytorch_model-00012-of-00015.bin",
628
  "model.layers.65.input_layernorm.weight": "pytorch_model-00012-of-00015.bin",
629
  "model.layers.65.mlp.down_proj.weight": "pytorch_model-00012-of-00015.bin",
@@ -633,7 +571,6 @@
633
  "model.layers.65.self_attn.k_proj.weight": "pytorch_model-00012-of-00015.bin",
634
  "model.layers.65.self_attn.o_proj.weight": "pytorch_model-00012-of-00015.bin",
635
  "model.layers.65.self_attn.q_proj.weight": "pytorch_model-00012-of-00015.bin",
636
- "model.layers.65.self_attn.rotary_emb.inv_freq": "pytorch_model-00012-of-00015.bin",
637
  "model.layers.65.self_attn.v_proj.weight": "pytorch_model-00012-of-00015.bin",
638
  "model.layers.66.input_layernorm.weight": "pytorch_model-00012-of-00015.bin",
639
  "model.layers.66.mlp.down_proj.weight": "pytorch_model-00012-of-00015.bin",
@@ -643,7 +580,6 @@
643
  "model.layers.66.self_attn.k_proj.weight": "pytorch_model-00012-of-00015.bin",
644
  "model.layers.66.self_attn.o_proj.weight": "pytorch_model-00012-of-00015.bin",
645
  "model.layers.66.self_attn.q_proj.weight": "pytorch_model-00012-of-00015.bin",
646
- "model.layers.66.self_attn.rotary_emb.inv_freq": "pytorch_model-00012-of-00015.bin",
647
  "model.layers.66.self_attn.v_proj.weight": "pytorch_model-00012-of-00015.bin",
648
  "model.layers.67.input_layernorm.weight": "pytorch_model-00012-of-00015.bin",
649
  "model.layers.67.mlp.down_proj.weight": "pytorch_model-00012-of-00015.bin",
@@ -653,7 +589,6 @@
653
  "model.layers.67.self_attn.k_proj.weight": "pytorch_model-00012-of-00015.bin",
654
  "model.layers.67.self_attn.o_proj.weight": "pytorch_model-00012-of-00015.bin",
655
  "model.layers.67.self_attn.q_proj.weight": "pytorch_model-00012-of-00015.bin",
656
- "model.layers.67.self_attn.rotary_emb.inv_freq": "pytorch_model-00012-of-00015.bin",
657
  "model.layers.67.self_attn.v_proj.weight": "pytorch_model-00012-of-00015.bin",
658
  "model.layers.68.input_layernorm.weight": "pytorch_model-00013-of-00015.bin",
659
  "model.layers.68.mlp.down_proj.weight": "pytorch_model-00013-of-00015.bin",
@@ -663,7 +598,6 @@
663
  "model.layers.68.self_attn.k_proj.weight": "pytorch_model-00012-of-00015.bin",
664
  "model.layers.68.self_attn.o_proj.weight": "pytorch_model-00012-of-00015.bin",
665
  "model.layers.68.self_attn.q_proj.weight": "pytorch_model-00012-of-00015.bin",
666
- "model.layers.68.self_attn.rotary_emb.inv_freq": "pytorch_model-00012-of-00015.bin",
667
  "model.layers.68.self_attn.v_proj.weight": "pytorch_model-00012-of-00015.bin",
668
  "model.layers.69.input_layernorm.weight": "pytorch_model-00013-of-00015.bin",
669
  "model.layers.69.mlp.down_proj.weight": "pytorch_model-00013-of-00015.bin",
@@ -673,7 +607,6 @@
673
  "model.layers.69.self_attn.k_proj.weight": "pytorch_model-00013-of-00015.bin",
674
  "model.layers.69.self_attn.o_proj.weight": "pytorch_model-00013-of-00015.bin",
675
  "model.layers.69.self_attn.q_proj.weight": "pytorch_model-00013-of-00015.bin",
676
- "model.layers.69.self_attn.rotary_emb.inv_freq": "pytorch_model-00013-of-00015.bin",
677
  "model.layers.69.self_attn.v_proj.weight": "pytorch_model-00013-of-00015.bin",
678
  "model.layers.7.input_layernorm.weight": "pytorch_model-00002-of-00015.bin",
679
  "model.layers.7.mlp.down_proj.weight": "pytorch_model-00002-of-00015.bin",
@@ -683,7 +616,6 @@
683
  "model.layers.7.self_attn.k_proj.weight": "pytorch_model-00002-of-00015.bin",
684
  "model.layers.7.self_attn.o_proj.weight": "pytorch_model-00002-of-00015.bin",
685
  "model.layers.7.self_attn.q_proj.weight": "pytorch_model-00002-of-00015.bin",
686
- "model.layers.7.self_attn.rotary_emb.inv_freq": "pytorch_model-00002-of-00015.bin",
687
  "model.layers.7.self_attn.v_proj.weight": "pytorch_model-00002-of-00015.bin",
688
  "model.layers.70.input_layernorm.weight": "pytorch_model-00013-of-00015.bin",
689
  "model.layers.70.mlp.down_proj.weight": "pytorch_model-00013-of-00015.bin",
@@ -693,7 +625,6 @@
693
  "model.layers.70.self_attn.k_proj.weight": "pytorch_model-00013-of-00015.bin",
694
  "model.layers.70.self_attn.o_proj.weight": "pytorch_model-00013-of-00015.bin",
695
  "model.layers.70.self_attn.q_proj.weight": "pytorch_model-00013-of-00015.bin",
696
- "model.layers.70.self_attn.rotary_emb.inv_freq": "pytorch_model-00013-of-00015.bin",
697
  "model.layers.70.self_attn.v_proj.weight": "pytorch_model-00013-of-00015.bin",
698
  "model.layers.71.input_layernorm.weight": "pytorch_model-00013-of-00015.bin",
699
  "model.layers.71.mlp.down_proj.weight": "pytorch_model-00013-of-00015.bin",
@@ -703,7 +634,6 @@
703
  "model.layers.71.self_attn.k_proj.weight": "pytorch_model-00013-of-00015.bin",
704
  "model.layers.71.self_attn.o_proj.weight": "pytorch_model-00013-of-00015.bin",
705
  "model.layers.71.self_attn.q_proj.weight": "pytorch_model-00013-of-00015.bin",
706
- "model.layers.71.self_attn.rotary_emb.inv_freq": "pytorch_model-00013-of-00015.bin",
707
  "model.layers.71.self_attn.v_proj.weight": "pytorch_model-00013-of-00015.bin",
708
  "model.layers.72.input_layernorm.weight": "pytorch_model-00013-of-00015.bin",
709
  "model.layers.72.mlp.down_proj.weight": "pytorch_model-00013-of-00015.bin",
@@ -713,7 +643,6 @@
713
  "model.layers.72.self_attn.k_proj.weight": "pytorch_model-00013-of-00015.bin",
714
  "model.layers.72.self_attn.o_proj.weight": "pytorch_model-00013-of-00015.bin",
715
  "model.layers.72.self_attn.q_proj.weight": "pytorch_model-00013-of-00015.bin",
716
- "model.layers.72.self_attn.rotary_emb.inv_freq": "pytorch_model-00013-of-00015.bin",
717
  "model.layers.72.self_attn.v_proj.weight": "pytorch_model-00013-of-00015.bin",
718
  "model.layers.73.input_layernorm.weight": "pytorch_model-00013-of-00015.bin",
719
  "model.layers.73.mlp.down_proj.weight": "pytorch_model-00013-of-00015.bin",
@@ -723,7 +652,6 @@
723
  "model.layers.73.self_attn.k_proj.weight": "pytorch_model-00013-of-00015.bin",
724
  "model.layers.73.self_attn.o_proj.weight": "pytorch_model-00013-of-00015.bin",
725
  "model.layers.73.self_attn.q_proj.weight": "pytorch_model-00013-of-00015.bin",
726
- "model.layers.73.self_attn.rotary_emb.inv_freq": "pytorch_model-00013-of-00015.bin",
727
  "model.layers.73.self_attn.v_proj.weight": "pytorch_model-00013-of-00015.bin",
728
  "model.layers.74.input_layernorm.weight": "pytorch_model-00014-of-00015.bin",
729
  "model.layers.74.mlp.down_proj.weight": "pytorch_model-00014-of-00015.bin",
@@ -733,7 +661,6 @@
733
  "model.layers.74.self_attn.k_proj.weight": "pytorch_model-00013-of-00015.bin",
734
  "model.layers.74.self_attn.o_proj.weight": "pytorch_model-00013-of-00015.bin",
735
  "model.layers.74.self_attn.q_proj.weight": "pytorch_model-00013-of-00015.bin",
736
- "model.layers.74.self_attn.rotary_emb.inv_freq": "pytorch_model-00013-of-00015.bin",
737
  "model.layers.74.self_attn.v_proj.weight": "pytorch_model-00013-of-00015.bin",
738
  "model.layers.75.input_layernorm.weight": "pytorch_model-00014-of-00015.bin",
739
  "model.layers.75.mlp.down_proj.weight": "pytorch_model-00014-of-00015.bin",
@@ -743,7 +670,6 @@
743
  "model.layers.75.self_attn.k_proj.weight": "pytorch_model-00014-of-00015.bin",
744
  "model.layers.75.self_attn.o_proj.weight": "pytorch_model-00014-of-00015.bin",
745
  "model.layers.75.self_attn.q_proj.weight": "pytorch_model-00014-of-00015.bin",
746
- "model.layers.75.self_attn.rotary_emb.inv_freq": "pytorch_model-00014-of-00015.bin",
747
  "model.layers.75.self_attn.v_proj.weight": "pytorch_model-00014-of-00015.bin",
748
  "model.layers.76.input_layernorm.weight": "pytorch_model-00014-of-00015.bin",
749
  "model.layers.76.mlp.down_proj.weight": "pytorch_model-00014-of-00015.bin",
@@ -753,7 +679,6 @@
753
  "model.layers.76.self_attn.k_proj.weight": "pytorch_model-00014-of-00015.bin",
754
  "model.layers.76.self_attn.o_proj.weight": "pytorch_model-00014-of-00015.bin",
755
  "model.layers.76.self_attn.q_proj.weight": "pytorch_model-00014-of-00015.bin",
756
- "model.layers.76.self_attn.rotary_emb.inv_freq": "pytorch_model-00014-of-00015.bin",
757
  "model.layers.76.self_attn.v_proj.weight": "pytorch_model-00014-of-00015.bin",
758
  "model.layers.77.input_layernorm.weight": "pytorch_model-00014-of-00015.bin",
759
  "model.layers.77.mlp.down_proj.weight": "pytorch_model-00014-of-00015.bin",
@@ -763,7 +688,6 @@
763
  "model.layers.77.self_attn.k_proj.weight": "pytorch_model-00014-of-00015.bin",
764
  "model.layers.77.self_attn.o_proj.weight": "pytorch_model-00014-of-00015.bin",
765
  "model.layers.77.self_attn.q_proj.weight": "pytorch_model-00014-of-00015.bin",
766
- "model.layers.77.self_attn.rotary_emb.inv_freq": "pytorch_model-00014-of-00015.bin",
767
  "model.layers.77.self_attn.v_proj.weight": "pytorch_model-00014-of-00015.bin",
768
  "model.layers.78.input_layernorm.weight": "pytorch_model-00014-of-00015.bin",
769
  "model.layers.78.mlp.down_proj.weight": "pytorch_model-00014-of-00015.bin",
@@ -773,7 +697,6 @@
773
  "model.layers.78.self_attn.k_proj.weight": "pytorch_model-00014-of-00015.bin",
774
  "model.layers.78.self_attn.o_proj.weight": "pytorch_model-00014-of-00015.bin",
775
  "model.layers.78.self_attn.q_proj.weight": "pytorch_model-00014-of-00015.bin",
776
- "model.layers.78.self_attn.rotary_emb.inv_freq": "pytorch_model-00014-of-00015.bin",
777
  "model.layers.78.self_attn.v_proj.weight": "pytorch_model-00014-of-00015.bin",
778
  "model.layers.79.input_layernorm.weight": "pytorch_model-00014-of-00015.bin",
779
  "model.layers.79.mlp.down_proj.weight": "pytorch_model-00014-of-00015.bin",
@@ -783,7 +706,6 @@
783
  "model.layers.79.self_attn.k_proj.weight": "pytorch_model-00014-of-00015.bin",
784
  "model.layers.79.self_attn.o_proj.weight": "pytorch_model-00014-of-00015.bin",
785
  "model.layers.79.self_attn.q_proj.weight": "pytorch_model-00014-of-00015.bin",
786
- "model.layers.79.self_attn.rotary_emb.inv_freq": "pytorch_model-00014-of-00015.bin",
787
  "model.layers.79.self_attn.v_proj.weight": "pytorch_model-00014-of-00015.bin",
788
  "model.layers.8.input_layernorm.weight": "pytorch_model-00002-of-00015.bin",
789
  "model.layers.8.mlp.down_proj.weight": "pytorch_model-00002-of-00015.bin",
@@ -793,7 +715,6 @@
793
  "model.layers.8.self_attn.k_proj.weight": "pytorch_model-00002-of-00015.bin",
794
  "model.layers.8.self_attn.o_proj.weight": "pytorch_model-00002-of-00015.bin",
795
  "model.layers.8.self_attn.q_proj.weight": "pytorch_model-00002-of-00015.bin",
796
- "model.layers.8.self_attn.rotary_emb.inv_freq": "pytorch_model-00002-of-00015.bin",
797
  "model.layers.8.self_attn.v_proj.weight": "pytorch_model-00002-of-00015.bin",
798
  "model.layers.9.input_layernorm.weight": "pytorch_model-00002-of-00015.bin",
799
  "model.layers.9.mlp.down_proj.weight": "pytorch_model-00002-of-00015.bin",
@@ -803,7 +724,6 @@
803
  "model.layers.9.self_attn.k_proj.weight": "pytorch_model-00002-of-00015.bin",
804
  "model.layers.9.self_attn.o_proj.weight": "pytorch_model-00002-of-00015.bin",
805
  "model.layers.9.self_attn.q_proj.weight": "pytorch_model-00002-of-00015.bin",
806
- "model.layers.9.self_attn.rotary_emb.inv_freq": "pytorch_model-00002-of-00015.bin",
807
  "model.layers.9.self_attn.v_proj.weight": "pytorch_model-00002-of-00015.bin",
808
  "model.norm.weight": "pytorch_model-00014-of-00015.bin"
809
  }
 
1
  {
2
  "metadata": {
3
+ "total_size": 137957490688
4
  },
5
  "weight_map": {
6
  "lm_head.weight": "pytorch_model-00015-of-00015.bin",
 
13
  "model.layers.0.self_attn.k_proj.weight": "pytorch_model-00001-of-00015.bin",
14
  "model.layers.0.self_attn.o_proj.weight": "pytorch_model-00001-of-00015.bin",
15
  "model.layers.0.self_attn.q_proj.weight": "pytorch_model-00001-of-00015.bin",
 
16
  "model.layers.0.self_attn.v_proj.weight": "pytorch_model-00001-of-00015.bin",
17
  "model.layers.1.input_layernorm.weight": "pytorch_model-00001-of-00015.bin",
18
  "model.layers.1.mlp.down_proj.weight": "pytorch_model-00001-of-00015.bin",
 
22
  "model.layers.1.self_attn.k_proj.weight": "pytorch_model-00001-of-00015.bin",
23
  "model.layers.1.self_attn.o_proj.weight": "pytorch_model-00001-of-00015.bin",
24
  "model.layers.1.self_attn.q_proj.weight": "pytorch_model-00001-of-00015.bin",
 
25
  "model.layers.1.self_attn.v_proj.weight": "pytorch_model-00001-of-00015.bin",
26
  "model.layers.10.input_layernorm.weight": "pytorch_model-00002-of-00015.bin",
27
  "model.layers.10.mlp.down_proj.weight": "pytorch_model-00002-of-00015.bin",
 
31
  "model.layers.10.self_attn.k_proj.weight": "pytorch_model-00002-of-00015.bin",
32
  "model.layers.10.self_attn.o_proj.weight": "pytorch_model-00002-of-00015.bin",
33
  "model.layers.10.self_attn.q_proj.weight": "pytorch_model-00002-of-00015.bin",
 
34
  "model.layers.10.self_attn.v_proj.weight": "pytorch_model-00002-of-00015.bin",
35
  "model.layers.11.input_layernorm.weight": "pytorch_model-00003-of-00015.bin",
36
  "model.layers.11.mlp.down_proj.weight": "pytorch_model-00003-of-00015.bin",
 
40
  "model.layers.11.self_attn.k_proj.weight": "pytorch_model-00002-of-00015.bin",
41
  "model.layers.11.self_attn.o_proj.weight": "pytorch_model-00002-of-00015.bin",
42
  "model.layers.11.self_attn.q_proj.weight": "pytorch_model-00002-of-00015.bin",
 
43
  "model.layers.11.self_attn.v_proj.weight": "pytorch_model-00002-of-00015.bin",
44
  "model.layers.12.input_layernorm.weight": "pytorch_model-00003-of-00015.bin",
45
  "model.layers.12.mlp.down_proj.weight": "pytorch_model-00003-of-00015.bin",
 
49
  "model.layers.12.self_attn.k_proj.weight": "pytorch_model-00003-of-00015.bin",
50
  "model.layers.12.self_attn.o_proj.weight": "pytorch_model-00003-of-00015.bin",
51
  "model.layers.12.self_attn.q_proj.weight": "pytorch_model-00003-of-00015.bin",
 
52
  "model.layers.12.self_attn.v_proj.weight": "pytorch_model-00003-of-00015.bin",
53
  "model.layers.13.input_layernorm.weight": "pytorch_model-00003-of-00015.bin",
54
  "model.layers.13.mlp.down_proj.weight": "pytorch_model-00003-of-00015.bin",
 
58
  "model.layers.13.self_attn.k_proj.weight": "pytorch_model-00003-of-00015.bin",
59
  "model.layers.13.self_attn.o_proj.weight": "pytorch_model-00003-of-00015.bin",
60
  "model.layers.13.self_attn.q_proj.weight": "pytorch_model-00003-of-00015.bin",
 
61
  "model.layers.13.self_attn.v_proj.weight": "pytorch_model-00003-of-00015.bin",
62
  "model.layers.14.input_layernorm.weight": "pytorch_model-00003-of-00015.bin",
63
  "model.layers.14.mlp.down_proj.weight": "pytorch_model-00003-of-00015.bin",
 
67
  "model.layers.14.self_attn.k_proj.weight": "pytorch_model-00003-of-00015.bin",
68
  "model.layers.14.self_attn.o_proj.weight": "pytorch_model-00003-of-00015.bin",
69
  "model.layers.14.self_attn.q_proj.weight": "pytorch_model-00003-of-00015.bin",
 
70
  "model.layers.14.self_attn.v_proj.weight": "pytorch_model-00003-of-00015.bin",
71
  "model.layers.15.input_layernorm.weight": "pytorch_model-00003-of-00015.bin",
72
  "model.layers.15.mlp.down_proj.weight": "pytorch_model-00003-of-00015.bin",
 
76
  "model.layers.15.self_attn.k_proj.weight": "pytorch_model-00003-of-00015.bin",
77
  "model.layers.15.self_attn.o_proj.weight": "pytorch_model-00003-of-00015.bin",
78
  "model.layers.15.self_attn.q_proj.weight": "pytorch_model-00003-of-00015.bin",
 
79
  "model.layers.15.self_attn.v_proj.weight": "pytorch_model-00003-of-00015.bin",
80
  "model.layers.16.input_layernorm.weight": "pytorch_model-00003-of-00015.bin",
81
  "model.layers.16.mlp.down_proj.weight": "pytorch_model-00003-of-00015.bin",
 
85
  "model.layers.16.self_attn.k_proj.weight": "pytorch_model-00003-of-00015.bin",
86
  "model.layers.16.self_attn.o_proj.weight": "pytorch_model-00003-of-00015.bin",
87
  "model.layers.16.self_attn.q_proj.weight": "pytorch_model-00003-of-00015.bin",
 
88
  "model.layers.16.self_attn.v_proj.weight": "pytorch_model-00003-of-00015.bin",
89
  "model.layers.17.input_layernorm.weight": "pytorch_model-00004-of-00015.bin",
90
  "model.layers.17.mlp.down_proj.weight": "pytorch_model-00004-of-00015.bin",
 
94
  "model.layers.17.self_attn.k_proj.weight": "pytorch_model-00004-of-00015.bin",
95
  "model.layers.17.self_attn.o_proj.weight": "pytorch_model-00004-of-00015.bin",
96
  "model.layers.17.self_attn.q_proj.weight": "pytorch_model-00004-of-00015.bin",
 
97
  "model.layers.17.self_attn.v_proj.weight": "pytorch_model-00004-of-00015.bin",
98
  "model.layers.18.input_layernorm.weight": "pytorch_model-00004-of-00015.bin",
99
  "model.layers.18.mlp.down_proj.weight": "pytorch_model-00004-of-00015.bin",
 
103
  "model.layers.18.self_attn.k_proj.weight": "pytorch_model-00004-of-00015.bin",
104
  "model.layers.18.self_attn.o_proj.weight": "pytorch_model-00004-of-00015.bin",
105
  "model.layers.18.self_attn.q_proj.weight": "pytorch_model-00004-of-00015.bin",
 
106
  "model.layers.18.self_attn.v_proj.weight": "pytorch_model-00004-of-00015.bin",
107
  "model.layers.19.input_layernorm.weight": "pytorch_model-00004-of-00015.bin",
108
  "model.layers.19.mlp.down_proj.weight": "pytorch_model-00004-of-00015.bin",
 
112
  "model.layers.19.self_attn.k_proj.weight": "pytorch_model-00004-of-00015.bin",
113
  "model.layers.19.self_attn.o_proj.weight": "pytorch_model-00004-of-00015.bin",
114
  "model.layers.19.self_attn.q_proj.weight": "pytorch_model-00004-of-00015.bin",
 
115
  "model.layers.19.self_attn.v_proj.weight": "pytorch_model-00004-of-00015.bin",
116
  "model.layers.2.input_layernorm.weight": "pytorch_model-00001-of-00015.bin",
117
  "model.layers.2.mlp.down_proj.weight": "pytorch_model-00001-of-00015.bin",
 
121
  "model.layers.2.self_attn.k_proj.weight": "pytorch_model-00001-of-00015.bin",
122
  "model.layers.2.self_attn.o_proj.weight": "pytorch_model-00001-of-00015.bin",
123
  "model.layers.2.self_attn.q_proj.weight": "pytorch_model-00001-of-00015.bin",
 
124
  "model.layers.2.self_attn.v_proj.weight": "pytorch_model-00001-of-00015.bin",
125
  "model.layers.20.input_layernorm.weight": "pytorch_model-00004-of-00015.bin",
126
  "model.layers.20.mlp.down_proj.weight": "pytorch_model-00004-of-00015.bin",
 
130
  "model.layers.20.self_attn.k_proj.weight": "pytorch_model-00004-of-00015.bin",
131
  "model.layers.20.self_attn.o_proj.weight": "pytorch_model-00004-of-00015.bin",
132
  "model.layers.20.self_attn.q_proj.weight": "pytorch_model-00004-of-00015.bin",
 
133
  "model.layers.20.self_attn.v_proj.weight": "pytorch_model-00004-of-00015.bin",
134
  "model.layers.21.input_layernorm.weight": "pytorch_model-00004-of-00015.bin",
135
  "model.layers.21.mlp.down_proj.weight": "pytorch_model-00004-of-00015.bin",
 
139
  "model.layers.21.self_attn.k_proj.weight": "pytorch_model-00004-of-00015.bin",
140
  "model.layers.21.self_attn.o_proj.weight": "pytorch_model-00004-of-00015.bin",
141
  "model.layers.21.self_attn.q_proj.weight": "pytorch_model-00004-of-00015.bin",
 
142
  "model.layers.21.self_attn.v_proj.weight": "pytorch_model-00004-of-00015.bin",
143
  "model.layers.22.input_layernorm.weight": "pytorch_model-00005-of-00015.bin",
144
  "model.layers.22.mlp.down_proj.weight": "pytorch_model-00005-of-00015.bin",
 
148
  "model.layers.22.self_attn.k_proj.weight": "pytorch_model-00004-of-00015.bin",
149
  "model.layers.22.self_attn.o_proj.weight": "pytorch_model-00004-of-00015.bin",
150
  "model.layers.22.self_attn.q_proj.weight": "pytorch_model-00004-of-00015.bin",
 
151
  "model.layers.22.self_attn.v_proj.weight": "pytorch_model-00004-of-00015.bin",
152
  "model.layers.23.input_layernorm.weight": "pytorch_model-00005-of-00015.bin",
153
  "model.layers.23.mlp.down_proj.weight": "pytorch_model-00005-of-00015.bin",
 
157
  "model.layers.23.self_attn.k_proj.weight": "pytorch_model-00005-of-00015.bin",
158
  "model.layers.23.self_attn.o_proj.weight": "pytorch_model-00005-of-00015.bin",
159
  "model.layers.23.self_attn.q_proj.weight": "pytorch_model-00005-of-00015.bin",
 
160
  "model.layers.23.self_attn.v_proj.weight": "pytorch_model-00005-of-00015.bin",
161
  "model.layers.24.input_layernorm.weight": "pytorch_model-00005-of-00015.bin",
162
  "model.layers.24.mlp.down_proj.weight": "pytorch_model-00005-of-00015.bin",
 
166
  "model.layers.24.self_attn.k_proj.weight": "pytorch_model-00005-of-00015.bin",
167
  "model.layers.24.self_attn.o_proj.weight": "pytorch_model-00005-of-00015.bin",
168
  "model.layers.24.self_attn.q_proj.weight": "pytorch_model-00005-of-00015.bin",
 
169
  "model.layers.24.self_attn.v_proj.weight": "pytorch_model-00005-of-00015.bin",
170
  "model.layers.25.input_layernorm.weight": "pytorch_model-00005-of-00015.bin",
171
  "model.layers.25.mlp.down_proj.weight": "pytorch_model-00005-of-00015.bin",
 
175
  "model.layers.25.self_attn.k_proj.weight": "pytorch_model-00005-of-00015.bin",
176
  "model.layers.25.self_attn.o_proj.weight": "pytorch_model-00005-of-00015.bin",
177
  "model.layers.25.self_attn.q_proj.weight": "pytorch_model-00005-of-00015.bin",
 
178
  "model.layers.25.self_attn.v_proj.weight": "pytorch_model-00005-of-00015.bin",
179
  "model.layers.26.input_layernorm.weight": "pytorch_model-00005-of-00015.bin",
180
  "model.layers.26.mlp.down_proj.weight": "pytorch_model-00005-of-00015.bin",
 
184
  "model.layers.26.self_attn.k_proj.weight": "pytorch_model-00005-of-00015.bin",
185
  "model.layers.26.self_attn.o_proj.weight": "pytorch_model-00005-of-00015.bin",
186
  "model.layers.26.self_attn.q_proj.weight": "pytorch_model-00005-of-00015.bin",
 
187
  "model.layers.26.self_attn.v_proj.weight": "pytorch_model-00005-of-00015.bin",
188
  "model.layers.27.input_layernorm.weight": "pytorch_model-00005-of-00015.bin",
189
  "model.layers.27.mlp.down_proj.weight": "pytorch_model-00005-of-00015.bin",
 
193
  "model.layers.27.self_attn.k_proj.weight": "pytorch_model-00005-of-00015.bin",
194
  "model.layers.27.self_attn.o_proj.weight": "pytorch_model-00005-of-00015.bin",
195
  "model.layers.27.self_attn.q_proj.weight": "pytorch_model-00005-of-00015.bin",
 
196
  "model.layers.27.self_attn.v_proj.weight": "pytorch_model-00005-of-00015.bin",
197
  "model.layers.28.input_layernorm.weight": "pytorch_model-00006-of-00015.bin",
198
  "model.layers.28.mlp.down_proj.weight": "pytorch_model-00006-of-00015.bin",
 
202
  "model.layers.28.self_attn.k_proj.weight": "pytorch_model-00005-of-00015.bin",
203
  "model.layers.28.self_attn.o_proj.weight": "pytorch_model-00005-of-00015.bin",
204
  "model.layers.28.self_attn.q_proj.weight": "pytorch_model-00005-of-00015.bin",
 
205
  "model.layers.28.self_attn.v_proj.weight": "pytorch_model-00005-of-00015.bin",
206
  "model.layers.29.input_layernorm.weight": "pytorch_model-00006-of-00015.bin",
207
  "model.layers.29.mlp.down_proj.weight": "pytorch_model-00006-of-00015.bin",
 
211
  "model.layers.29.self_attn.k_proj.weight": "pytorch_model-00006-of-00015.bin",
212
  "model.layers.29.self_attn.o_proj.weight": "pytorch_model-00006-of-00015.bin",
213
  "model.layers.29.self_attn.q_proj.weight": "pytorch_model-00006-of-00015.bin",
 
214
  "model.layers.29.self_attn.v_proj.weight": "pytorch_model-00006-of-00015.bin",
215
  "model.layers.3.input_layernorm.weight": "pytorch_model-00001-of-00015.bin",
216
  "model.layers.3.mlp.down_proj.weight": "pytorch_model-00001-of-00015.bin",
 
220
  "model.layers.3.self_attn.k_proj.weight": "pytorch_model-00001-of-00015.bin",
221
  "model.layers.3.self_attn.o_proj.weight": "pytorch_model-00001-of-00015.bin",
222
  "model.layers.3.self_attn.q_proj.weight": "pytorch_model-00001-of-00015.bin",
 
223
  "model.layers.3.self_attn.v_proj.weight": "pytorch_model-00001-of-00015.bin",
224
  "model.layers.30.input_layernorm.weight": "pytorch_model-00006-of-00015.bin",
225
  "model.layers.30.mlp.down_proj.weight": "pytorch_model-00006-of-00015.bin",
 
229
  "model.layers.30.self_attn.k_proj.weight": "pytorch_model-00006-of-00015.bin",
230
  "model.layers.30.self_attn.o_proj.weight": "pytorch_model-00006-of-00015.bin",
231
  "model.layers.30.self_attn.q_proj.weight": "pytorch_model-00006-of-00015.bin",
 
232
  "model.layers.30.self_attn.v_proj.weight": "pytorch_model-00006-of-00015.bin",
233
  "model.layers.31.input_layernorm.weight": "pytorch_model-00006-of-00015.bin",
234
  "model.layers.31.mlp.down_proj.weight": "pytorch_model-00006-of-00015.bin",
 
238
  "model.layers.31.self_attn.k_proj.weight": "pytorch_model-00006-of-00015.bin",
239
  "model.layers.31.self_attn.o_proj.weight": "pytorch_model-00006-of-00015.bin",
240
  "model.layers.31.self_attn.q_proj.weight": "pytorch_model-00006-of-00015.bin",
 
241
  "model.layers.31.self_attn.v_proj.weight": "pytorch_model-00006-of-00015.bin",
242
  "model.layers.32.input_layernorm.weight": "pytorch_model-00006-of-00015.bin",
243
  "model.layers.32.mlp.down_proj.weight": "pytorch_model-00006-of-00015.bin",
 
247
  "model.layers.32.self_attn.k_proj.weight": "pytorch_model-00006-of-00015.bin",
248
  "model.layers.32.self_attn.o_proj.weight": "pytorch_model-00006-of-00015.bin",
249
  "model.layers.32.self_attn.q_proj.weight": "pytorch_model-00006-of-00015.bin",
 
250
  "model.layers.32.self_attn.v_proj.weight": "pytorch_model-00006-of-00015.bin",
251
  "model.layers.33.input_layernorm.weight": "pytorch_model-00006-of-00015.bin",
252
  "model.layers.33.mlp.down_proj.weight": "pytorch_model-00006-of-00015.bin",
 
256
  "model.layers.33.self_attn.k_proj.weight": "pytorch_model-00006-of-00015.bin",
257
  "model.layers.33.self_attn.o_proj.weight": "pytorch_model-00006-of-00015.bin",
258
  "model.layers.33.self_attn.q_proj.weight": "pytorch_model-00006-of-00015.bin",
 
259
  "model.layers.33.self_attn.v_proj.weight": "pytorch_model-00006-of-00015.bin",
260
  "model.layers.34.input_layernorm.weight": "pytorch_model-00007-of-00015.bin",
261
  "model.layers.34.mlp.down_proj.weight": "pytorch_model-00007-of-00015.bin",
 
265
  "model.layers.34.self_attn.k_proj.weight": "pytorch_model-00006-of-00015.bin",
266
  "model.layers.34.self_attn.o_proj.weight": "pytorch_model-00006-of-00015.bin",
267
  "model.layers.34.self_attn.q_proj.weight": "pytorch_model-00006-of-00015.bin",
 
268
  "model.layers.34.self_attn.v_proj.weight": "pytorch_model-00006-of-00015.bin",
269
  "model.layers.35.input_layernorm.weight": "pytorch_model-00007-of-00015.bin",
270
  "model.layers.35.mlp.down_proj.weight": "pytorch_model-00007-of-00015.bin",
 
274
  "model.layers.35.self_attn.k_proj.weight": "pytorch_model-00007-of-00015.bin",
275
  "model.layers.35.self_attn.o_proj.weight": "pytorch_model-00007-of-00015.bin",
276
  "model.layers.35.self_attn.q_proj.weight": "pytorch_model-00007-of-00015.bin",
 
277
  "model.layers.35.self_attn.v_proj.weight": "pytorch_model-00007-of-00015.bin",
278
  "model.layers.36.input_layernorm.weight": "pytorch_model-00007-of-00015.bin",
279
  "model.layers.36.mlp.down_proj.weight": "pytorch_model-00007-of-00015.bin",
 
283
  "model.layers.36.self_attn.k_proj.weight": "pytorch_model-00007-of-00015.bin",
284
  "model.layers.36.self_attn.o_proj.weight": "pytorch_model-00007-of-00015.bin",
285
  "model.layers.36.self_attn.q_proj.weight": "pytorch_model-00007-of-00015.bin",
 
286
  "model.layers.36.self_attn.v_proj.weight": "pytorch_model-00007-of-00015.bin",
287
  "model.layers.37.input_layernorm.weight": "pytorch_model-00007-of-00015.bin",
288
  "model.layers.37.mlp.down_proj.weight": "pytorch_model-00007-of-00015.bin",
 
292
  "model.layers.37.self_attn.k_proj.weight": "pytorch_model-00007-of-00015.bin",
293
  "model.layers.37.self_attn.o_proj.weight": "pytorch_model-00007-of-00015.bin",
294
  "model.layers.37.self_attn.q_proj.weight": "pytorch_model-00007-of-00015.bin",
 
295
  "model.layers.37.self_attn.v_proj.weight": "pytorch_model-00007-of-00015.bin",
296
  "model.layers.38.input_layernorm.weight": "pytorch_model-00007-of-00015.bin",
297
  "model.layers.38.mlp.down_proj.weight": "pytorch_model-00007-of-00015.bin",
 
301
  "model.layers.38.self_attn.k_proj.weight": "pytorch_model-00007-of-00015.bin",
302
  "model.layers.38.self_attn.o_proj.weight": "pytorch_model-00007-of-00015.bin",
303
  "model.layers.38.self_attn.q_proj.weight": "pytorch_model-00007-of-00015.bin",
 
304
  "model.layers.38.self_attn.v_proj.weight": "pytorch_model-00007-of-00015.bin",
305
  "model.layers.39.input_layernorm.weight": "pytorch_model-00007-of-00015.bin",
306
  "model.layers.39.mlp.down_proj.weight": "pytorch_model-00007-of-00015.bin",
 
310
  "model.layers.39.self_attn.k_proj.weight": "pytorch_model-00007-of-00015.bin",
311
  "model.layers.39.self_attn.o_proj.weight": "pytorch_model-00007-of-00015.bin",
312
  "model.layers.39.self_attn.q_proj.weight": "pytorch_model-00007-of-00015.bin",
 
313
  "model.layers.39.self_attn.v_proj.weight": "pytorch_model-00007-of-00015.bin",
314
  "model.layers.4.input_layernorm.weight": "pytorch_model-00001-of-00015.bin",
315
  "model.layers.4.mlp.down_proj.weight": "pytorch_model-00001-of-00015.bin",
 
319
  "model.layers.4.self_attn.k_proj.weight": "pytorch_model-00001-of-00015.bin",
320
  "model.layers.4.self_attn.o_proj.weight": "pytorch_model-00001-of-00015.bin",
321
  "model.layers.4.self_attn.q_proj.weight": "pytorch_model-00001-of-00015.bin",
 
322
  "model.layers.4.self_attn.v_proj.weight": "pytorch_model-00001-of-00015.bin",
323
  "model.layers.40.input_layernorm.weight": "pytorch_model-00008-of-00015.bin",
324
  "model.layers.40.mlp.down_proj.weight": "pytorch_model-00008-of-00015.bin",
 
328
  "model.layers.40.self_attn.k_proj.weight": "pytorch_model-00008-of-00015.bin",
329
  "model.layers.40.self_attn.o_proj.weight": "pytorch_model-00008-of-00015.bin",
330
  "model.layers.40.self_attn.q_proj.weight": "pytorch_model-00008-of-00015.bin",
 
331
  "model.layers.40.self_attn.v_proj.weight": "pytorch_model-00008-of-00015.bin",
332
  "model.layers.41.input_layernorm.weight": "pytorch_model-00008-of-00015.bin",
333
  "model.layers.41.mlp.down_proj.weight": "pytorch_model-00008-of-00015.bin",
 
337
  "model.layers.41.self_attn.k_proj.weight": "pytorch_model-00008-of-00015.bin",
338
  "model.layers.41.self_attn.o_proj.weight": "pytorch_model-00008-of-00015.bin",
339
  "model.layers.41.self_attn.q_proj.weight": "pytorch_model-00008-of-00015.bin",
 
340
  "model.layers.41.self_attn.v_proj.weight": "pytorch_model-00008-of-00015.bin",
341
  "model.layers.42.input_layernorm.weight": "pytorch_model-00008-of-00015.bin",
342
  "model.layers.42.mlp.down_proj.weight": "pytorch_model-00008-of-00015.bin",
 
346
  "model.layers.42.self_attn.k_proj.weight": "pytorch_model-00008-of-00015.bin",
347
  "model.layers.42.self_attn.o_proj.weight": "pytorch_model-00008-of-00015.bin",
348
  "model.layers.42.self_attn.q_proj.weight": "pytorch_model-00008-of-00015.bin",
 
349
  "model.layers.42.self_attn.v_proj.weight": "pytorch_model-00008-of-00015.bin",
350
  "model.layers.43.input_layernorm.weight": "pytorch_model-00008-of-00015.bin",
351
  "model.layers.43.mlp.down_proj.weight": "pytorch_model-00008-of-00015.bin",
 
355
  "model.layers.43.self_attn.k_proj.weight": "pytorch_model-00008-of-00015.bin",
356
  "model.layers.43.self_attn.o_proj.weight": "pytorch_model-00008-of-00015.bin",
357
  "model.layers.43.self_attn.q_proj.weight": "pytorch_model-00008-of-00015.bin",
 
358
  "model.layers.43.self_attn.v_proj.weight": "pytorch_model-00008-of-00015.bin",
359
  "model.layers.44.input_layernorm.weight": "pytorch_model-00008-of-00015.bin",
360
  "model.layers.44.mlp.down_proj.weight": "pytorch_model-00008-of-00015.bin",
 
364
  "model.layers.44.self_attn.k_proj.weight": "pytorch_model-00008-of-00015.bin",
365
  "model.layers.44.self_attn.o_proj.weight": "pytorch_model-00008-of-00015.bin",
366
  "model.layers.44.self_attn.q_proj.weight": "pytorch_model-00008-of-00015.bin",
 
367
  "model.layers.44.self_attn.v_proj.weight": "pytorch_model-00008-of-00015.bin",
368
  "model.layers.45.input_layernorm.weight": "pytorch_model-00009-of-00015.bin",
369
  "model.layers.45.mlp.down_proj.weight": "pytorch_model-00009-of-00015.bin",
 
373
  "model.layers.45.self_attn.k_proj.weight": "pytorch_model-00008-of-00015.bin",
374
  "model.layers.45.self_attn.o_proj.weight": "pytorch_model-00008-of-00015.bin",
375
  "model.layers.45.self_attn.q_proj.weight": "pytorch_model-00008-of-00015.bin",
 
376
  "model.layers.45.self_attn.v_proj.weight": "pytorch_model-00008-of-00015.bin",
377
  "model.layers.46.input_layernorm.weight": "pytorch_model-00009-of-00015.bin",
378
  "model.layers.46.mlp.down_proj.weight": "pytorch_model-00009-of-00015.bin",
 
382
  "model.layers.46.self_attn.k_proj.weight": "pytorch_model-00009-of-00015.bin",
383
  "model.layers.46.self_attn.o_proj.weight": "pytorch_model-00009-of-00015.bin",
384
  "model.layers.46.self_attn.q_proj.weight": "pytorch_model-00009-of-00015.bin",
 
385
  "model.layers.46.self_attn.v_proj.weight": "pytorch_model-00009-of-00015.bin",
386
  "model.layers.47.input_layernorm.weight": "pytorch_model-00009-of-00015.bin",
387
  "model.layers.47.mlp.down_proj.weight": "pytorch_model-00009-of-00015.bin",
 
391
  "model.layers.47.self_attn.k_proj.weight": "pytorch_model-00009-of-00015.bin",
392
  "model.layers.47.self_attn.o_proj.weight": "pytorch_model-00009-of-00015.bin",
393
  "model.layers.47.self_attn.q_proj.weight": "pytorch_model-00009-of-00015.bin",
 
394
  "model.layers.47.self_attn.v_proj.weight": "pytorch_model-00009-of-00015.bin",
395
  "model.layers.48.input_layernorm.weight": "pytorch_model-00009-of-00015.bin",
396
  "model.layers.48.mlp.down_proj.weight": "pytorch_model-00009-of-00015.bin",
 
400
  "model.layers.48.self_attn.k_proj.weight": "pytorch_model-00009-of-00015.bin",
401
  "model.layers.48.self_attn.o_proj.weight": "pytorch_model-00009-of-00015.bin",
402
  "model.layers.48.self_attn.q_proj.weight": "pytorch_model-00009-of-00015.bin",
 
403
  "model.layers.48.self_attn.v_proj.weight": "pytorch_model-00009-of-00015.bin",
404
  "model.layers.49.input_layernorm.weight": "pytorch_model-00009-of-00015.bin",
405
  "model.layers.49.mlp.down_proj.weight": "pytorch_model-00009-of-00015.bin",
 
409
  "model.layers.49.self_attn.k_proj.weight": "pytorch_model-00009-of-00015.bin",
410
  "model.layers.49.self_attn.o_proj.weight": "pytorch_model-00009-of-00015.bin",
411
  "model.layers.49.self_attn.q_proj.weight": "pytorch_model-00009-of-00015.bin",
 
412
  "model.layers.49.self_attn.v_proj.weight": "pytorch_model-00009-of-00015.bin",
413
  "model.layers.5.input_layernorm.weight": "pytorch_model-00002-of-00015.bin",
414
  "model.layers.5.mlp.down_proj.weight": "pytorch_model-00002-of-00015.bin",
 
418
  "model.layers.5.self_attn.k_proj.weight": "pytorch_model-00001-of-00015.bin",
419
  "model.layers.5.self_attn.o_proj.weight": "pytorch_model-00001-of-00015.bin",
420
  "model.layers.5.self_attn.q_proj.weight": "pytorch_model-00001-of-00015.bin",
 
421
  "model.layers.5.self_attn.v_proj.weight": "pytorch_model-00001-of-00015.bin",
422
  "model.layers.50.input_layernorm.weight": "pytorch_model-00009-of-00015.bin",
423
  "model.layers.50.mlp.down_proj.weight": "pytorch_model-00009-of-00015.bin",
 
427
  "model.layers.50.self_attn.k_proj.weight": "pytorch_model-00009-of-00015.bin",
428
  "model.layers.50.self_attn.o_proj.weight": "pytorch_model-00009-of-00015.bin",
429
  "model.layers.50.self_attn.q_proj.weight": "pytorch_model-00009-of-00015.bin",
 
430
  "model.layers.50.self_attn.v_proj.weight": "pytorch_model-00009-of-00015.bin",
431
  "model.layers.51.input_layernorm.weight": "pytorch_model-00010-of-00015.bin",
432
  "model.layers.51.mlp.down_proj.weight": "pytorch_model-00010-of-00015.bin",
 
436
  "model.layers.51.self_attn.k_proj.weight": "pytorch_model-00009-of-00015.bin",
437
  "model.layers.51.self_attn.o_proj.weight": "pytorch_model-00009-of-00015.bin",
438
  "model.layers.51.self_attn.q_proj.weight": "pytorch_model-00009-of-00015.bin",
 
439
  "model.layers.51.self_attn.v_proj.weight": "pytorch_model-00009-of-00015.bin",
440
  "model.layers.52.input_layernorm.weight": "pytorch_model-00010-of-00015.bin",
441
  "model.layers.52.mlp.down_proj.weight": "pytorch_model-00010-of-00015.bin",
 
445
  "model.layers.52.self_attn.k_proj.weight": "pytorch_model-00010-of-00015.bin",
446
  "model.layers.52.self_attn.o_proj.weight": "pytorch_model-00010-of-00015.bin",
447
  "model.layers.52.self_attn.q_proj.weight": "pytorch_model-00010-of-00015.bin",
 
448
  "model.layers.52.self_attn.v_proj.weight": "pytorch_model-00010-of-00015.bin",
449
  "model.layers.53.input_layernorm.weight": "pytorch_model-00010-of-00015.bin",
450
  "model.layers.53.mlp.down_proj.weight": "pytorch_model-00010-of-00015.bin",
 
454
  "model.layers.53.self_attn.k_proj.weight": "pytorch_model-00010-of-00015.bin",
455
  "model.layers.53.self_attn.o_proj.weight": "pytorch_model-00010-of-00015.bin",
456
  "model.layers.53.self_attn.q_proj.weight": "pytorch_model-00010-of-00015.bin",
 
457
  "model.layers.53.self_attn.v_proj.weight": "pytorch_model-00010-of-00015.bin",
458
  "model.layers.54.input_layernorm.weight": "pytorch_model-00010-of-00015.bin",
459
  "model.layers.54.mlp.down_proj.weight": "pytorch_model-00010-of-00015.bin",
 
463
  "model.layers.54.self_attn.k_proj.weight": "pytorch_model-00010-of-00015.bin",
464
  "model.layers.54.self_attn.o_proj.weight": "pytorch_model-00010-of-00015.bin",
465
  "model.layers.54.self_attn.q_proj.weight": "pytorch_model-00010-of-00015.bin",
 
466
  "model.layers.54.self_attn.v_proj.weight": "pytorch_model-00010-of-00015.bin",
467
  "model.layers.55.input_layernorm.weight": "pytorch_model-00010-of-00015.bin",
468
  "model.layers.55.mlp.down_proj.weight": "pytorch_model-00010-of-00015.bin",
 
472
  "model.layers.55.self_attn.k_proj.weight": "pytorch_model-00010-of-00015.bin",
473
  "model.layers.55.self_attn.o_proj.weight": "pytorch_model-00010-of-00015.bin",
474
  "model.layers.55.self_attn.q_proj.weight": "pytorch_model-00010-of-00015.bin",
 
475
  "model.layers.55.self_attn.v_proj.weight": "pytorch_model-00010-of-00015.bin",
476
  "model.layers.56.input_layernorm.weight": "pytorch_model-00010-of-00015.bin",
477
  "model.layers.56.mlp.down_proj.weight": "pytorch_model-00010-of-00015.bin",
 
481
  "model.layers.56.self_attn.k_proj.weight": "pytorch_model-00010-of-00015.bin",
482
  "model.layers.56.self_attn.o_proj.weight": "pytorch_model-00010-of-00015.bin",
483
  "model.layers.56.self_attn.q_proj.weight": "pytorch_model-00010-of-00015.bin",
 
484
  "model.layers.56.self_attn.v_proj.weight": "pytorch_model-00010-of-00015.bin",
485
  "model.layers.57.input_layernorm.weight": "pytorch_model-00011-of-00015.bin",
486
  "model.layers.57.mlp.down_proj.weight": "pytorch_model-00011-of-00015.bin",
 
490
  "model.layers.57.self_attn.k_proj.weight": "pytorch_model-00010-of-00015.bin",
491
  "model.layers.57.self_attn.o_proj.weight": "pytorch_model-00010-of-00015.bin",
492
  "model.layers.57.self_attn.q_proj.weight": "pytorch_model-00010-of-00015.bin",
 
493
  "model.layers.57.self_attn.v_proj.weight": "pytorch_model-00010-of-00015.bin",
494
  "model.layers.58.input_layernorm.weight": "pytorch_model-00011-of-00015.bin",
495
  "model.layers.58.mlp.down_proj.weight": "pytorch_model-00011-of-00015.bin",
 
499
  "model.layers.58.self_attn.k_proj.weight": "pytorch_model-00011-of-00015.bin",
500
  "model.layers.58.self_attn.o_proj.weight": "pytorch_model-00011-of-00015.bin",
501
  "model.layers.58.self_attn.q_proj.weight": "pytorch_model-00011-of-00015.bin",
 
502
  "model.layers.58.self_attn.v_proj.weight": "pytorch_model-00011-of-00015.bin",
503
  "model.layers.59.input_layernorm.weight": "pytorch_model-00011-of-00015.bin",
504
  "model.layers.59.mlp.down_proj.weight": "pytorch_model-00011-of-00015.bin",
 
508
  "model.layers.59.self_attn.k_proj.weight": "pytorch_model-00011-of-00015.bin",
509
  "model.layers.59.self_attn.o_proj.weight": "pytorch_model-00011-of-00015.bin",
510
  "model.layers.59.self_attn.q_proj.weight": "pytorch_model-00011-of-00015.bin",
 
511
  "model.layers.59.self_attn.v_proj.weight": "pytorch_model-00011-of-00015.bin",
512
  "model.layers.6.input_layernorm.weight": "pytorch_model-00002-of-00015.bin",
513
  "model.layers.6.mlp.down_proj.weight": "pytorch_model-00002-of-00015.bin",
 
517
  "model.layers.6.self_attn.k_proj.weight": "pytorch_model-00002-of-00015.bin",
518
  "model.layers.6.self_attn.o_proj.weight": "pytorch_model-00002-of-00015.bin",
519
  "model.layers.6.self_attn.q_proj.weight": "pytorch_model-00002-of-00015.bin",
 
520
  "model.layers.6.self_attn.v_proj.weight": "pytorch_model-00002-of-00015.bin",
521
  "model.layers.60.input_layernorm.weight": "pytorch_model-00011-of-00015.bin",
522
  "model.layers.60.mlp.down_proj.weight": "pytorch_model-00011-of-00015.bin",
 
526
  "model.layers.60.self_attn.k_proj.weight": "pytorch_model-00011-of-00015.bin",
527
  "model.layers.60.self_attn.o_proj.weight": "pytorch_model-00011-of-00015.bin",
528
  "model.layers.60.self_attn.q_proj.weight": "pytorch_model-00011-of-00015.bin",
 
529
  "model.layers.60.self_attn.v_proj.weight": "pytorch_model-00011-of-00015.bin",
530
  "model.layers.61.input_layernorm.weight": "pytorch_model-00011-of-00015.bin",
531
  "model.layers.61.mlp.down_proj.weight": "pytorch_model-00011-of-00015.bin",
 
535
  "model.layers.61.self_attn.k_proj.weight": "pytorch_model-00011-of-00015.bin",
536
  "model.layers.61.self_attn.o_proj.weight": "pytorch_model-00011-of-00015.bin",
537
  "model.layers.61.self_attn.q_proj.weight": "pytorch_model-00011-of-00015.bin",
 
538
  "model.layers.61.self_attn.v_proj.weight": "pytorch_model-00011-of-00015.bin",
539
  "model.layers.62.input_layernorm.weight": "pytorch_model-00011-of-00015.bin",
540
  "model.layers.62.mlp.down_proj.weight": "pytorch_model-00011-of-00015.bin",
 
544
  "model.layers.62.self_attn.k_proj.weight": "pytorch_model-00011-of-00015.bin",
545
  "model.layers.62.self_attn.o_proj.weight": "pytorch_model-00011-of-00015.bin",
546
  "model.layers.62.self_attn.q_proj.weight": "pytorch_model-00011-of-00015.bin",
 
547
  "model.layers.62.self_attn.v_proj.weight": "pytorch_model-00011-of-00015.bin",
548
  "model.layers.63.input_layernorm.weight": "pytorch_model-00012-of-00015.bin",
549
  "model.layers.63.mlp.down_proj.weight": "pytorch_model-00012-of-00015.bin",
 
553
  "model.layers.63.self_attn.k_proj.weight": "pytorch_model-00012-of-00015.bin",
554
  "model.layers.63.self_attn.o_proj.weight": "pytorch_model-00012-of-00015.bin",
555
  "model.layers.63.self_attn.q_proj.weight": "pytorch_model-00012-of-00015.bin",
 
556
  "model.layers.63.self_attn.v_proj.weight": "pytorch_model-00012-of-00015.bin",
557
  "model.layers.64.input_layernorm.weight": "pytorch_model-00012-of-00015.bin",
558
  "model.layers.64.mlp.down_proj.weight": "pytorch_model-00012-of-00015.bin",
 
562
  "model.layers.64.self_attn.k_proj.weight": "pytorch_model-00012-of-00015.bin",
563
  "model.layers.64.self_attn.o_proj.weight": "pytorch_model-00012-of-00015.bin",
564
  "model.layers.64.self_attn.q_proj.weight": "pytorch_model-00012-of-00015.bin",
 
565
  "model.layers.64.self_attn.v_proj.weight": "pytorch_model-00012-of-00015.bin",
566
  "model.layers.65.input_layernorm.weight": "pytorch_model-00012-of-00015.bin",
567
  "model.layers.65.mlp.down_proj.weight": "pytorch_model-00012-of-00015.bin",
 
571
  "model.layers.65.self_attn.k_proj.weight": "pytorch_model-00012-of-00015.bin",
572
  "model.layers.65.self_attn.o_proj.weight": "pytorch_model-00012-of-00015.bin",
573
  "model.layers.65.self_attn.q_proj.weight": "pytorch_model-00012-of-00015.bin",
 
574
  "model.layers.65.self_attn.v_proj.weight": "pytorch_model-00012-of-00015.bin",
575
  "model.layers.66.input_layernorm.weight": "pytorch_model-00012-of-00015.bin",
576
  "model.layers.66.mlp.down_proj.weight": "pytorch_model-00012-of-00015.bin",
 
580
  "model.layers.66.self_attn.k_proj.weight": "pytorch_model-00012-of-00015.bin",
581
  "model.layers.66.self_attn.o_proj.weight": "pytorch_model-00012-of-00015.bin",
582
  "model.layers.66.self_attn.q_proj.weight": "pytorch_model-00012-of-00015.bin",
 
583
  "model.layers.66.self_attn.v_proj.weight": "pytorch_model-00012-of-00015.bin",
584
  "model.layers.67.input_layernorm.weight": "pytorch_model-00012-of-00015.bin",
585
  "model.layers.67.mlp.down_proj.weight": "pytorch_model-00012-of-00015.bin",
 
589
  "model.layers.67.self_attn.k_proj.weight": "pytorch_model-00012-of-00015.bin",
590
  "model.layers.67.self_attn.o_proj.weight": "pytorch_model-00012-of-00015.bin",
591
  "model.layers.67.self_attn.q_proj.weight": "pytorch_model-00012-of-00015.bin",
 
592
  "model.layers.67.self_attn.v_proj.weight": "pytorch_model-00012-of-00015.bin",
593
  "model.layers.68.input_layernorm.weight": "pytorch_model-00013-of-00015.bin",
594
  "model.layers.68.mlp.down_proj.weight": "pytorch_model-00013-of-00015.bin",
 
598
  "model.layers.68.self_attn.k_proj.weight": "pytorch_model-00012-of-00015.bin",
599
  "model.layers.68.self_attn.o_proj.weight": "pytorch_model-00012-of-00015.bin",
600
  "model.layers.68.self_attn.q_proj.weight": "pytorch_model-00012-of-00015.bin",
 
601
  "model.layers.68.self_attn.v_proj.weight": "pytorch_model-00012-of-00015.bin",
602
  "model.layers.69.input_layernorm.weight": "pytorch_model-00013-of-00015.bin",
603
  "model.layers.69.mlp.down_proj.weight": "pytorch_model-00013-of-00015.bin",
 
607
  "model.layers.69.self_attn.k_proj.weight": "pytorch_model-00013-of-00015.bin",
608
  "model.layers.69.self_attn.o_proj.weight": "pytorch_model-00013-of-00015.bin",
609
  "model.layers.69.self_attn.q_proj.weight": "pytorch_model-00013-of-00015.bin",
 
610
  "model.layers.69.self_attn.v_proj.weight": "pytorch_model-00013-of-00015.bin",
611
  "model.layers.7.input_layernorm.weight": "pytorch_model-00002-of-00015.bin",
612
  "model.layers.7.mlp.down_proj.weight": "pytorch_model-00002-of-00015.bin",
 
616
  "model.layers.7.self_attn.k_proj.weight": "pytorch_model-00002-of-00015.bin",
617
  "model.layers.7.self_attn.o_proj.weight": "pytorch_model-00002-of-00015.bin",
618
  "model.layers.7.self_attn.q_proj.weight": "pytorch_model-00002-of-00015.bin",
 
619
  "model.layers.7.self_attn.v_proj.weight": "pytorch_model-00002-of-00015.bin",
620
  "model.layers.70.input_layernorm.weight": "pytorch_model-00013-of-00015.bin",
621
  "model.layers.70.mlp.down_proj.weight": "pytorch_model-00013-of-00015.bin",
 
625
  "model.layers.70.self_attn.k_proj.weight": "pytorch_model-00013-of-00015.bin",
626
  "model.layers.70.self_attn.o_proj.weight": "pytorch_model-00013-of-00015.bin",
627
  "model.layers.70.self_attn.q_proj.weight": "pytorch_model-00013-of-00015.bin",
 
628
  "model.layers.70.self_attn.v_proj.weight": "pytorch_model-00013-of-00015.bin",
629
  "model.layers.71.input_layernorm.weight": "pytorch_model-00013-of-00015.bin",
630
  "model.layers.71.mlp.down_proj.weight": "pytorch_model-00013-of-00015.bin",
 
634
  "model.layers.71.self_attn.k_proj.weight": "pytorch_model-00013-of-00015.bin",
635
  "model.layers.71.self_attn.o_proj.weight": "pytorch_model-00013-of-00015.bin",
636
  "model.layers.71.self_attn.q_proj.weight": "pytorch_model-00013-of-00015.bin",
 
637
  "model.layers.71.self_attn.v_proj.weight": "pytorch_model-00013-of-00015.bin",
638
  "model.layers.72.input_layernorm.weight": "pytorch_model-00013-of-00015.bin",
639
  "model.layers.72.mlp.down_proj.weight": "pytorch_model-00013-of-00015.bin",
 
643
  "model.layers.72.self_attn.k_proj.weight": "pytorch_model-00013-of-00015.bin",
644
  "model.layers.72.self_attn.o_proj.weight": "pytorch_model-00013-of-00015.bin",
645
  "model.layers.72.self_attn.q_proj.weight": "pytorch_model-00013-of-00015.bin",
 
646
  "model.layers.72.self_attn.v_proj.weight": "pytorch_model-00013-of-00015.bin",
647
  "model.layers.73.input_layernorm.weight": "pytorch_model-00013-of-00015.bin",
648
  "model.layers.73.mlp.down_proj.weight": "pytorch_model-00013-of-00015.bin",
 
652
  "model.layers.73.self_attn.k_proj.weight": "pytorch_model-00013-of-00015.bin",
653
  "model.layers.73.self_attn.o_proj.weight": "pytorch_model-00013-of-00015.bin",
654
  "model.layers.73.self_attn.q_proj.weight": "pytorch_model-00013-of-00015.bin",
 
655
  "model.layers.73.self_attn.v_proj.weight": "pytorch_model-00013-of-00015.bin",
656
  "model.layers.74.input_layernorm.weight": "pytorch_model-00014-of-00015.bin",
657
  "model.layers.74.mlp.down_proj.weight": "pytorch_model-00014-of-00015.bin",
 
661
  "model.layers.74.self_attn.k_proj.weight": "pytorch_model-00013-of-00015.bin",
662
  "model.layers.74.self_attn.o_proj.weight": "pytorch_model-00013-of-00015.bin",
663
  "model.layers.74.self_attn.q_proj.weight": "pytorch_model-00013-of-00015.bin",
 
664
  "model.layers.74.self_attn.v_proj.weight": "pytorch_model-00013-of-00015.bin",
665
  "model.layers.75.input_layernorm.weight": "pytorch_model-00014-of-00015.bin",
666
  "model.layers.75.mlp.down_proj.weight": "pytorch_model-00014-of-00015.bin",
 
670
  "model.layers.75.self_attn.k_proj.weight": "pytorch_model-00014-of-00015.bin",
671
  "model.layers.75.self_attn.o_proj.weight": "pytorch_model-00014-of-00015.bin",
672
  "model.layers.75.self_attn.q_proj.weight": "pytorch_model-00014-of-00015.bin",
 
673
  "model.layers.75.self_attn.v_proj.weight": "pytorch_model-00014-of-00015.bin",
674
  "model.layers.76.input_layernorm.weight": "pytorch_model-00014-of-00015.bin",
675
  "model.layers.76.mlp.down_proj.weight": "pytorch_model-00014-of-00015.bin",
 
679
  "model.layers.76.self_attn.k_proj.weight": "pytorch_model-00014-of-00015.bin",
680
  "model.layers.76.self_attn.o_proj.weight": "pytorch_model-00014-of-00015.bin",
681
  "model.layers.76.self_attn.q_proj.weight": "pytorch_model-00014-of-00015.bin",
 
682
  "model.layers.76.self_attn.v_proj.weight": "pytorch_model-00014-of-00015.bin",
683
  "model.layers.77.input_layernorm.weight": "pytorch_model-00014-of-00015.bin",
684
  "model.layers.77.mlp.down_proj.weight": "pytorch_model-00014-of-00015.bin",
 
688
  "model.layers.77.self_attn.k_proj.weight": "pytorch_model-00014-of-00015.bin",
689
  "model.layers.77.self_attn.o_proj.weight": "pytorch_model-00014-of-00015.bin",
690
  "model.layers.77.self_attn.q_proj.weight": "pytorch_model-00014-of-00015.bin",
 
691
  "model.layers.77.self_attn.v_proj.weight": "pytorch_model-00014-of-00015.bin",
692
  "model.layers.78.input_layernorm.weight": "pytorch_model-00014-of-00015.bin",
693
  "model.layers.78.mlp.down_proj.weight": "pytorch_model-00014-of-00015.bin",
 
697
  "model.layers.78.self_attn.k_proj.weight": "pytorch_model-00014-of-00015.bin",
698
  "model.layers.78.self_attn.o_proj.weight": "pytorch_model-00014-of-00015.bin",
699
  "model.layers.78.self_attn.q_proj.weight": "pytorch_model-00014-of-00015.bin",
 
700
  "model.layers.78.self_attn.v_proj.weight": "pytorch_model-00014-of-00015.bin",
701
  "model.layers.79.input_layernorm.weight": "pytorch_model-00014-of-00015.bin",
702
  "model.layers.79.mlp.down_proj.weight": "pytorch_model-00014-of-00015.bin",
 
706
  "model.layers.79.self_attn.k_proj.weight": "pytorch_model-00014-of-00015.bin",
707
  "model.layers.79.self_attn.o_proj.weight": "pytorch_model-00014-of-00015.bin",
708
  "model.layers.79.self_attn.q_proj.weight": "pytorch_model-00014-of-00015.bin",
 
709
  "model.layers.79.self_attn.v_proj.weight": "pytorch_model-00014-of-00015.bin",
710
  "model.layers.8.input_layernorm.weight": "pytorch_model-00002-of-00015.bin",
711
  "model.layers.8.mlp.down_proj.weight": "pytorch_model-00002-of-00015.bin",
 
715
  "model.layers.8.self_attn.k_proj.weight": "pytorch_model-00002-of-00015.bin",
716
  "model.layers.8.self_attn.o_proj.weight": "pytorch_model-00002-of-00015.bin",
717
  "model.layers.8.self_attn.q_proj.weight": "pytorch_model-00002-of-00015.bin",
 
718
  "model.layers.8.self_attn.v_proj.weight": "pytorch_model-00002-of-00015.bin",
719
  "model.layers.9.input_layernorm.weight": "pytorch_model-00002-of-00015.bin",
720
  "model.layers.9.mlp.down_proj.weight": "pytorch_model-00002-of-00015.bin",
 
724
  "model.layers.9.self_attn.k_proj.weight": "pytorch_model-00002-of-00015.bin",
725
  "model.layers.9.self_attn.o_proj.weight": "pytorch_model-00002-of-00015.bin",
726
  "model.layers.9.self_attn.q_proj.weight": "pytorch_model-00002-of-00015.bin",
 
727
  "model.layers.9.self_attn.v_proj.weight": "pytorch_model-00002-of-00015.bin",
728
  "model.norm.weight": "pytorch_model-00014-of-00015.bin"
729
  }
tokenizer.json CHANGED
@@ -114,12 +114,6 @@
114
  "post_processor": {
115
  "type": "TemplateProcessing",
116
  "single": [
117
- {
118
- "SpecialToken": {
119
- "id": "<s>",
120
- "type_id": 0
121
- }
122
- },
123
  {
124
  "Sequence": {
125
  "id": "A",
@@ -128,24 +122,12 @@
128
  }
129
  ],
130
  "pair": [
131
- {
132
- "SpecialToken": {
133
- "id": "<s>",
134
- "type_id": 0
135
- }
136
- },
137
  {
138
  "Sequence": {
139
  "id": "A",
140
  "type_id": 0
141
  }
142
  },
143
- {
144
- "SpecialToken": {
145
- "id": "<s>",
146
- "type_id": 1
147
- }
148
- },
149
  {
150
  "Sequence": {
151
  "id": "B",
@@ -153,17 +135,7 @@
153
  }
154
  }
155
  ],
156
- "special_tokens": {
157
- "<s>": {
158
- "id": "<s>",
159
- "ids": [
160
- 1
161
- ],
162
- "tokens": [
163
- "<s>"
164
- ]
165
- }
166
- }
167
  },
168
  "decoder": {
169
  "type": "Sequence",
 
114
  "post_processor": {
115
  "type": "TemplateProcessing",
116
  "single": [
 
 
 
 
 
 
117
  {
118
  "Sequence": {
119
  "id": "A",
 
122
  }
123
  ],
124
  "pair": [
 
 
 
 
 
 
125
  {
126
  "Sequence": {
127
  "id": "A",
128
  "type_id": 0
129
  }
130
  },
 
 
 
 
 
 
131
  {
132
  "Sequence": {
133
  "id": "B",
 
135
  }
136
  }
137
  ],
138
+ "special_tokens": {}
 
 
 
 
 
 
 
 
 
 
139
  },
140
  "decoder": {
141
  "type": "Sequence",
tokenizer_config.json CHANGED
@@ -30,5 +30,6 @@
30
  "normalized": false,
31
  "rstrip": false,
32
  "single_word": false
33
- }
 
34
  }
 
30
  "normalized": false,
31
  "rstrip": false,
32
  "single_word": false
33
+ },
34
+ "use_default_system_prompt": false
35
  }