jondurbin commited on
Commit
75e3bc9
1 Parent(s): e1aa7e4

Upload folder using huggingface_hub

Browse files
config.json CHANGED
@@ -1,5 +1,5 @@
1
  {
2
- "_name_or_path": "airoboros-l2-70b-2.1",
3
  "architectures": [
4
  "LlamaForCausalLM"
5
  ],
@@ -16,11 +16,22 @@
16
  "num_key_value_heads": 8,
17
  "pad_token_id": 0,
18
  "pretraining_tp": 1,
 
 
 
 
 
 
 
 
 
 
 
19
  "rms_norm_eps": 1e-05,
20
  "rope_scaling": null,
21
  "tie_word_embeddings": false,
22
- "torch_dtype": "float16",
23
- "transformers_version": "4.32.0",
24
  "use_cache": true,
25
  "vocab_size": 32000
26
  }
 
1
  {
2
+ "_name_or_path": "llama-2-70b-hf",
3
  "architectures": [
4
  "LlamaForCausalLM"
5
  ],
 
16
  "num_key_value_heads": 8,
17
  "pad_token_id": 0,
18
  "pretraining_tp": 1,
19
+ "quantization_config": {
20
+ "bnb_4bit_compute_dtype": "bfloat16",
21
+ "bnb_4bit_quant_type": "nf4",
22
+ "bnb_4bit_use_double_quant": true,
23
+ "llm_int8_enable_fp32_cpu_offload": false,
24
+ "llm_int8_has_fp16_weight": false,
25
+ "llm_int8_skip_modules": null,
26
+ "llm_int8_threshold": 6.0,
27
+ "load_in_4bit": true,
28
+ "load_in_8bit": false
29
+ },
30
  "rms_norm_eps": 1e-05,
31
  "rope_scaling": null,
32
  "tie_word_embeddings": false,
33
+ "torch_dtype": "bfloat16",
34
+ "transformers_version": "4.31.0",
35
  "use_cache": true,
36
  "vocab_size": 32000
37
  }
generation_config.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 1,
4
+ "eos_token_id": 2,
5
+ "pad_token_id": 0,
6
+ "transformers_version": "4.31.0"
7
+ }
pytorch_model-00001-of-00015.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:74a2f40ebf50eaefc9dae2aa112aa33c208673abfad817d6bffcc75e799af2f8
3
- size 9852602021
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ee58e2decaf06e2685bddfb5ef6237ba4a57db70291b3485e95746754a08e86b
3
+ size 9852604917
pytorch_model-00002-of-00015.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:71aa3e3c2df3f94e430a7cf70195f6f9b07582fc4b0419f3c12492262e426c52
3
- size 9798109597
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8a0d736aab9f511c20040c9225cc1b90b4e689a8ccbd3d5a85015eca4d76975c
3
+ size 9798112505
pytorch_model-00003-of-00015.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:f44759ea6d4a839c71f315a4fb7947630d9f0d3e8a0ec47a86d749ccd7d5e904
3
- size 9965880719
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:08c54d2f0e4d5f4c68e83662b4a10a1b6fa1b08a51c8f3dd8e0f157e42254db8
3
+ size 9965883221
pytorch_model-00004-of-00015.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:2aff0c62769cd76495dc4a627506ddb0dba2dab14075706594804dd2e7357981
3
- size 9798076069
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:348de6b2462fb4583dd9eda16d3ff715e7049e83da7a46c5243cfc47e7bfce3c
3
+ size 9798079017
pytorch_model-00005-of-00015.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:fdd51f8227b72edf1acb2e6192b2ca5214f16bc6d0050bd20c3719db34182243
3
- size 9798109577
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:af21f6085bf5ee2438e836b557652a58671eeb5f3444b00a178e2f0780c70a23
3
+ size 9798112545
pytorch_model-00006-of-00015.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:43c150ea4831df8a1e3e64f3b85f333442cfc85515d611dda8179e89b8be44cb
3
- size 9798109661
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fc80f5ec49959c1100d822af69d3f6fb7252a9d4266ff383b7bd9b912434fecf
3
+ size 9798112569
pytorch_model-00007-of-00015.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c39567c084c0c87cc488efde168e0079de4689b7c64a20941064ee8bb98e48ff
3
- size 9965880719
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f510910a62c24a0e38e1f56e98f617dfe99bf0df6a2887be481e5aea3d839daf
3
+ size 9965883221
pytorch_model-00008-of-00015.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:9bc7708b13e790f519f1c9aefe55fd2bb687d89493104ce7c4f22b08c3805338
3
- size 9798076069
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d552432c4b18da2cdcd99759fac0cbd4921fe335a9ea264a299cb6bc953bec71
3
+ size 9798079017
pytorch_model-00009-of-00015.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:4b69c8cc44a197a3f9453ff75b83a852f48c75580e075dd2f5bbbd67e733e89f
3
- size 9798109577
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e68a9f26ee80a8109b1600bbdc62ea89bfb26c2dd05c9d4b5ef6ddee0621ad00
3
+ size 9798112545
pytorch_model-00010-of-00015.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:89078dae1a7a7bf462049edd4c564198820f9b931522726d36669b72cf97e737
3
- size 9798109661
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:30f54ed6675acc669f08c186e798a9e9afa4d433d05ef2bddd44be6a1e5d2cf0
3
+ size 9798112569
pytorch_model-00011-of-00015.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c874fd8987a15487aa68e70973883bb5676152285faef64c87767e2daa0c5b3c
3
- size 9965880719
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:241fff5006a0b9263e8126d364c1de39aa0816e0728273205119bb7721f8472d
3
+ size 9965883221
pytorch_model-00012-of-00015.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:04750765fceac920f25e67b6f89b88233d51632d6f02cb6bf7f5131308da60f5
3
- size 9798076069
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f1e5d9f190afe76221ca19be6d6fa05a0125b3c907a88d3c0589611c142d7298
3
+ size 9798079017
pytorch_model-00013-of-00015.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:831a541270726ff826c2bbfeac80533a4f40b60661a7b5090e791a87d75dd46d
3
- size 9798109577
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:43d4a2667e4177f339aff0a7b8076219ee9a05f14044f12c4156c93da6552ef4
3
+ size 9798112545
pytorch_model-00014-of-00015.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:34ba5c968def14d48fbd0d387c9503ab33ab5597b1e691a4fa23b8ba3315abce
3
- size 9496135043
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6adfc71e15f3a1e5d41c0f3d001bd726374148dff310007fbd6790274f7bc134
3
+ size 9496137473
pytorch_model-00015-of-00015.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e57333b2ffb8bc8fd6147c737fb3d3422c2c80b6608c9f8d91e58ddca1b5431c
3
  size 524288938
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4affec791e4217a4e28d166b327c27a59e7b3693d7c5940d29e825a1dbbd1a17
3
  size 524288938
pytorch_model.bin.index.json CHANGED
@@ -1,6 +1,6 @@
1
  {
2
  "metadata": {
3
- "total_size": 137953296384
4
  },
5
  "weight_map": {
6
  "lm_head.weight": "pytorch_model-00015-of-00015.bin",
@@ -13,6 +13,7 @@
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",
@@ -22,6 +23,7 @@
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",
@@ -31,6 +33,7 @@
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",
@@ -40,6 +43,7 @@
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",
@@ -49,6 +53,7 @@
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",
@@ -58,6 +63,7 @@
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",
@@ -67,6 +73,7 @@
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",
@@ -76,6 +83,7 @@
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",
@@ -85,6 +93,7 @@
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",
@@ -94,6 +103,7 @@
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",
@@ -103,6 +113,7 @@
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",
@@ -112,6 +123,7 @@
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",
@@ -121,6 +133,7 @@
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",
@@ -130,6 +143,7 @@
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",
@@ -139,6 +153,7 @@
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",
@@ -148,6 +163,7 @@
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",
@@ -157,6 +173,7 @@
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",
@@ -166,6 +183,7 @@
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",
@@ -175,6 +193,7 @@
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",
@@ -184,6 +203,7 @@
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",
@@ -193,6 +213,7 @@
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",
@@ -202,6 +223,7 @@
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",
@@ -211,6 +233,7 @@
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",
@@ -220,6 +243,7 @@
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",
@@ -229,6 +253,7 @@
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",
@@ -238,6 +263,7 @@
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",
@@ -247,6 +273,7 @@
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",
@@ -256,6 +283,7 @@
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",
@@ -265,6 +293,7 @@
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",
@@ -274,6 +303,7 @@
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",
@@ -283,6 +313,7 @@
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",
@@ -292,6 +323,7 @@
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",
@@ -301,6 +333,7 @@
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",
@@ -310,6 +343,7 @@
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",
@@ -319,6 +353,7 @@
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",
@@ -328,6 +363,7 @@
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",
@@ -337,6 +373,7 @@
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",
@@ -346,6 +383,7 @@
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",
@@ -355,6 +393,7 @@
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",
@@ -364,6 +403,7 @@
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",
@@ -373,6 +413,7 @@
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",
@@ -382,6 +423,7 @@
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",
@@ -391,6 +433,7 @@
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",
@@ -400,6 +443,7 @@
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",
@@ -409,6 +453,7 @@
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",
@@ -418,6 +463,7 @@
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",
@@ -427,6 +473,7 @@
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",
@@ -436,6 +483,7 @@
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",
@@ -445,6 +493,7 @@
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",
@@ -454,6 +503,7 @@
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",
@@ -463,6 +513,7 @@
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",
@@ -472,6 +523,7 @@
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",
@@ -481,6 +533,7 @@
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",
@@ -490,6 +543,7 @@
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",
@@ -499,6 +553,7 @@
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",
@@ -508,6 +563,7 @@
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",
@@ -517,6 +573,7 @@
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",
@@ -526,6 +583,7 @@
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",
@@ -535,6 +593,7 @@
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",
@@ -544,6 +603,7 @@
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",
@@ -553,6 +613,7 @@
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",
@@ -562,6 +623,7 @@
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",
@@ -571,6 +633,7 @@
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",
@@ -580,6 +643,7 @@
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",
@@ -589,6 +653,7 @@
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",
@@ -598,6 +663,7 @@
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",
@@ -607,6 +673,7 @@
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",
@@ -616,6 +683,7 @@
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",
@@ -625,6 +693,7 @@
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",
@@ -634,6 +703,7 @@
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",
@@ -643,6 +713,7 @@
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",
@@ -652,6 +723,7 @@
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",
@@ -661,6 +733,7 @@
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",
@@ -670,6 +743,7 @@
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",
@@ -679,6 +753,7 @@
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",
@@ -688,6 +763,7 @@
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",
@@ -697,6 +773,7 @@
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",
@@ -706,6 +783,7 @@
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",
@@ -715,6 +793,7 @@
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",
@@ -724,6 +803,7 @@
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
  }
 
1
  {
2
  "metadata": {
3
+ "total_size": 137953306624
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.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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  "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
  }
special_tokens_map.json ADDED
@@ -0,0 +1,23 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "</s>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "unk_token": {
17
+ "content": "<unk>",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ }
23
+ }