Hugging Face's logo

fdschmidt93
/
NLLB-LLM2Vec-Meta-Llama-31-8B-Instruct-mntp-unsup-simcse

Sentence Similarity
Transformers
Safetensors
multilingual
nllb-llm2vec
feature-extraction
text-embedding
embeddings
information-retrieval
beir
text-classification
language-model
text-clustering
text-semantic-similarity
text-evaluation
text-reranking
Sentence Similarity
natural_questions
ms_marco
fever
hotpot_qa
mteb
custom_code
Model card Files Community
1
Fabian-David Schmidt commited on
Commit
ba88283
•
1 Parent(s): a2e7ca9

feat: update model weights

Browse files
Files changed (8) hide show
  1. model-00001-of-00007.safetensors → model-00001-of-00004.safetensors +2 -2
  2. model-00002-of-00007.safetensors → model-00002-of-00004.safetensors +2 -2
  3. model-00003-of-00007.safetensors → model-00003-of-00004.safetensors +2 -2
  4. model-00004-of-00007.safetensors → model-00004-of-00004.safetensors +2 -2
  5. model-00005-of-00007.safetensors +0 -3
  6. model-00006-of-00007.safetensors +0 -3
  7. model-00007-of-00007.safetensors +0 -3
  8. model.safetensors.index.json +487 -487
model-00001-of-00007.safetensors → model-00001-of-00004.safetensors RENAMED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e0e2d6673950aef18051379de3f8b9ac71079213a2396ed41eaad5f3325317fc
3
- size 4795637544
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ab7c10637f2b030eb90fa09642dbf6ac7279c96bf454a7378c8402542e085aae
3
+ size 4981626968
model-00002-of-00007.safetensors → model-00002-of-00004.safetensors RENAMED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c14b9b0619077a8a40c86cbce9d22d57e65ebf13ab4b847bf870846d66e748f4
3
- size 4832007544
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:39afe2dfd0d0ab2f4c2ba552a409f2c2a685cb57f1d9e0bf78863ffd6129198a
3
+ size 4949470904
model-00003-of-00007.safetensors → model-00003-of-00004.safetensors RENAMED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b03fd3bb595b5bce41d77c48b08d39fe6e064036e1ca684702c010cb7876f954
3
- size 4999813200
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:276dfcdc542f28f6c978a1921d4e876a502e3506acd69bffbff5078e5c6e9662
3
+ size 4915916376
model-00004-of-00007.safetensors → model-00004-of-00004.safetensors RENAMED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:67faaafa1cb268d84031a540951a4ae69f96c5b57b29562bedd9efe18c9f0a6a
3
- size 4999813232
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9a4c15e8a42992daf957a2dbe03ed21d819a9c5c145e5e44815fb575959a1c1b
3
+ size 998304352
model-00005-of-00007.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:4290ffc52ad6a3bfc4b652ce629ca08504d607a24ce2688ce17c6bfc6a09c56f
3
- size 4832007592
 
 
 
 
model-00006-of-00007.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:47b47969386a099ab2486cac4d892ccea975a6d97ccb3c603038b60d46592b7a
3
- size 4999813232
 
 
 
 
model-00007-of-00007.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:9ada5ca6cf7bd855bf6953c89e2bfb54b9b1de7829cb8ca48fb7139b49e5d63d
3
- size 2231487200
 
 
 
 
model.safetensors.index.json CHANGED
@@ -1,493 +1,493 @@
1
  {
2
  "metadata": {
3
- "total_size": 31690522624
4
  },
5
  "weight_map": {
6
- "llm2vec.embed_tokens.weight": "model-00001-of-00007.safetensors",
7
- "llm2vec.layers.0.input_layernorm.weight": "model-00001-of-00007.safetensors",
8
- "llm2vec.layers.0.mlp.down_proj.weight": "model-00001-of-00007.safetensors",
9
- "llm2vec.layers.0.mlp.gate_proj.weight": "model-00001-of-00007.safetensors",
10
- "llm2vec.layers.0.mlp.up_proj.weight": "model-00001-of-00007.safetensors",
11
- "llm2vec.layers.0.post_attention_layernorm.weight": "model-00001-of-00007.safetensors",
12
- "llm2vec.layers.0.self_attn.k_proj.weight": "model-00001-of-00007.safetensors",
13
- "llm2vec.layers.0.self_attn.o_proj.weight": "model-00001-of-00007.safetensors",
14
- "llm2vec.layers.0.self_attn.q_proj.weight": "model-00001-of-00007.safetensors",
15
- "llm2vec.layers.0.self_attn.v_proj.weight": "model-00001-of-00007.safetensors",
16
- "llm2vec.layers.1.input_layernorm.weight": "model-00002-of-00007.safetensors",
17
- "llm2vec.layers.1.mlp.down_proj.weight": "model-00002-of-00007.safetensors",
18
- "llm2vec.layers.1.mlp.gate_proj.weight": "model-00002-of-00007.safetensors",
19
- "llm2vec.layers.1.mlp.up_proj.weight": "model-00002-of-00007.safetensors",
20
- "llm2vec.layers.1.post_attention_layernorm.weight": "model-00002-of-00007.safetensors",
21
- "llm2vec.layers.1.self_attn.k_proj.weight": "model-00001-of-00007.safetensors",
22
- "llm2vec.layers.1.self_attn.o_proj.weight": "model-00001-of-00007.safetensors",
23
- "llm2vec.layers.1.self_attn.q_proj.weight": "model-00001-of-00007.safetensors",
24
- "llm2vec.layers.1.self_attn.v_proj.weight": "model-00001-of-00007.safetensors",
25
- "llm2vec.layers.10.input_layernorm.weight": "model-00003-of-00007.safetensors",
26
- "llm2vec.layers.10.mlp.down_proj.weight": "model-00003-of-00007.safetensors",
27
- "llm2vec.layers.10.mlp.gate_proj.weight": "model-00003-of-00007.safetensors",
28
- "llm2vec.layers.10.mlp.up_proj.weight": "model-00003-of-00007.safetensors",
29
- "llm2vec.layers.10.post_attention_layernorm.weight": "model-00003-of-00007.safetensors",
30
- "llm2vec.layers.10.self_attn.k_proj.weight": "model-00003-of-00007.safetensors",
31
- "llm2vec.layers.10.self_attn.o_proj.weight": "model-00003-of-00007.safetensors",
32
- "llm2vec.layers.10.self_attn.q_proj.weight": "model-00003-of-00007.safetensors",
33
- "llm2vec.layers.10.self_attn.v_proj.weight": "model-00003-of-00007.safetensors",
34
- "llm2vec.layers.11.input_layernorm.weight": "model-00003-of-00007.safetensors",
35
- "llm2vec.layers.11.mlp.down_proj.weight": "model-00003-of-00007.safetensors",
36
- "llm2vec.layers.11.mlp.gate_proj.weight": "model-00003-of-00007.safetensors",
37
- "llm2vec.layers.11.mlp.up_proj.weight": "model-00003-of-00007.safetensors",
38
- "llm2vec.layers.11.post_attention_layernorm.weight": "model-00003-of-00007.safetensors",
39
- "llm2vec.layers.11.self_attn.k_proj.weight": "model-00003-of-00007.safetensors",
40
- "llm2vec.layers.11.self_attn.o_proj.weight": "model-00003-of-00007.safetensors",
41
- "llm2vec.layers.11.self_attn.q_proj.weight": "model-00003-of-00007.safetensors",
42
- "llm2vec.layers.11.self_attn.v_proj.weight": "model-00003-of-00007.safetensors",
43
- "llm2vec.layers.12.input_layernorm.weight": "model-00004-of-00007.safetensors",
44
- "llm2vec.layers.12.mlp.down_proj.weight": "model-00004-of-00007.safetensors",
45
- "llm2vec.layers.12.mlp.gate_proj.weight": "model-00003-of-00007.safetensors",
46
- "llm2vec.layers.12.mlp.up_proj.weight": "model-00004-of-00007.safetensors",
47
- "llm2vec.layers.12.post_attention_layernorm.weight": "model-00004-of-00007.safetensors",
48
- "llm2vec.layers.12.self_attn.k_proj.weight": "model-00003-of-00007.safetensors",
49
- "llm2vec.layers.12.self_attn.o_proj.weight": "model-00003-of-00007.safetensors",
50
- "llm2vec.layers.12.self_attn.q_proj.weight": "model-00003-of-00007.safetensors",
51
- "llm2vec.layers.12.self_attn.v_proj.weight": "model-00003-of-00007.safetensors",
52
- "llm2vec.layers.13.input_layernorm.weight": "model-00004-of-00007.safetensors",
53
- "llm2vec.layers.13.mlp.down_proj.weight": "model-00004-of-00007.safetensors",
54
- "llm2vec.layers.13.mlp.gate_proj.weight": "model-00004-of-00007.safetensors",
55
- "llm2vec.layers.13.mlp.up_proj.weight": "model-00004-of-00007.safetensors",
56
- "llm2vec.layers.13.post_attention_layernorm.weight": "model-00004-of-00007.safetensors",
57
- "llm2vec.layers.13.self_attn.k_proj.weight": "model-00004-of-00007.safetensors",
58
- "llm2vec.layers.13.self_attn.o_proj.weight": "model-00004-of-00007.safetensors",
59
- "llm2vec.layers.13.self_attn.q_proj.weight": "model-00004-of-00007.safetensors",
60
- "llm2vec.layers.13.self_attn.v_proj.weight": "model-00004-of-00007.safetensors",
61
- "llm2vec.layers.14.input_layernorm.weight": "model-00004-of-00007.safetensors",
62
- "llm2vec.layers.14.mlp.down_proj.weight": "model-00004-of-00007.safetensors",
63
- "llm2vec.layers.14.mlp.gate_proj.weight": "model-00004-of-00007.safetensors",
64
- "llm2vec.layers.14.mlp.up_proj.weight": "model-00004-of-00007.safetensors",
65
- "llm2vec.layers.14.post_attention_layernorm.weight": "model-00004-of-00007.safetensors",
66
- "llm2vec.layers.14.self_attn.k_proj.weight": "model-00004-of-00007.safetensors",
67
- "llm2vec.layers.14.self_attn.o_proj.weight": "model-00004-of-00007.safetensors",
68
- "llm2vec.layers.14.self_attn.q_proj.weight": "model-00004-of-00007.safetensors",
69
- "llm2vec.layers.14.self_attn.v_proj.weight": "model-00004-of-00007.safetensors",
70
- "llm2vec.layers.15.input_layernorm.weight": "model-00004-of-00007.safetensors",
71
- "llm2vec.layers.15.mlp.down_proj.weight": "model-00004-of-00007.safetensors",
72
- "llm2vec.layers.15.mlp.gate_proj.weight": "model-00004-of-00007.safetensors",
73
- "llm2vec.layers.15.mlp.up_proj.weight": "model-00004-of-00007.safetensors",
74
- "llm2vec.layers.15.post_attention_layernorm.weight": "model-00004-of-00007.safetensors",
75
- "llm2vec.layers.15.self_attn.k_proj.weight": "model-00004-of-00007.safetensors",
76
- "llm2vec.layers.15.self_attn.o_proj.weight": "model-00004-of-00007.safetensors",
77
- "llm2vec.layers.15.self_attn.q_proj.weight": "model-00004-of-00007.safetensors",
78
- "llm2vec.layers.15.self_attn.v_proj.weight": "model-00004-of-00007.safetensors",
79
- "llm2vec.layers.16.input_layernorm.weight": "model-00004-of-00007.safetensors",
80
- "llm2vec.layers.16.mlp.down_proj.weight": "model-00004-of-00007.safetensors",
81
- "llm2vec.layers.16.mlp.gate_proj.weight": "model-00004-of-00007.safetensors",
82
- "llm2vec.layers.16.mlp.up_proj.weight": "model-00004-of-00007.safetensors",
83
- "llm2vec.layers.16.post_attention_layernorm.weight": "model-00004-of-00007.safetensors",
84
- "llm2vec.layers.16.self_attn.k_proj.weight": "model-00004-of-00007.safetensors",
85
- "llm2vec.layers.16.self_attn.o_proj.weight": "model-00004-of-00007.safetensors",
86
- "llm2vec.layers.16.self_attn.q_proj.weight": "model-00004-of-00007.safetensors",
87
- "llm2vec.layers.16.self_attn.v_proj.weight": "model-00004-of-00007.safetensors",
88
- "llm2vec.layers.17.input_layernorm.weight": "model-00004-of-00007.safetensors",
89
- "llm2vec.layers.17.mlp.down_proj.weight": "model-00004-of-00007.safetensors",
90
- "llm2vec.layers.17.mlp.gate_proj.weight": "model-00004-of-00007.safetensors",
91
- "llm2vec.layers.17.mlp.up_proj.weight": "model-00004-of-00007.safetensors",
92
- "llm2vec.layers.17.post_attention_layernorm.weight": "model-00004-of-00007.safetensors",
93
- "llm2vec.layers.17.self_attn.k_proj.weight": "model-00004-of-00007.safetensors",
94
- "llm2vec.layers.17.self_attn.o_proj.weight": "model-00004-of-00007.safetensors",
95
- "llm2vec.layers.17.self_attn.q_proj.weight": "model-00004-of-00007.safetensors",
96
- "llm2vec.layers.17.self_attn.v_proj.weight": "model-00004-of-00007.safetensors",
97
- "llm2vec.layers.18.input_layernorm.weight": "model-00005-of-00007.safetensors",
98
- "llm2vec.layers.18.mlp.down_proj.weight": "model-00005-of-00007.safetensors",
99
- "llm2vec.layers.18.mlp.gate_proj.weight": "model-00005-of-00007.safetensors",
100
- "llm2vec.layers.18.mlp.up_proj.weight": "model-00005-of-00007.safetensors",
101
- "llm2vec.layers.18.post_attention_layernorm.weight": "model-00005-of-00007.safetensors",
102
- "llm2vec.layers.18.self_attn.k_proj.weight": "model-00004-of-00007.safetensors",
103
- "llm2vec.layers.18.self_attn.o_proj.weight": "model-00004-of-00007.safetensors",
104
- "llm2vec.layers.18.self_attn.q_proj.weight": "model-00004-of-00007.safetensors",
105
- "llm2vec.layers.18.self_attn.v_proj.weight": "model-00004-of-00007.safetensors",
106
- "llm2vec.layers.19.input_layernorm.weight": "model-00005-of-00007.safetensors",
107
- "llm2vec.layers.19.mlp.down_proj.weight": "model-00005-of-00007.safetensors",
108
- "llm2vec.layers.19.mlp.gate_proj.weight": "model-00005-of-00007.safetensors",
109
- "llm2vec.layers.19.mlp.up_proj.weight": "model-00005-of-00007.safetensors",
110
- "llm2vec.layers.19.post_attention_layernorm.weight": "model-00005-of-00007.safetensors",
111
- "llm2vec.layers.19.self_attn.k_proj.weight": "model-00005-of-00007.safetensors",
112
- "llm2vec.layers.19.self_attn.o_proj.weight": "model-00005-of-00007.safetensors",
113
- "llm2vec.layers.19.self_attn.q_proj.weight": "model-00005-of-00007.safetensors",
114
- "llm2vec.layers.19.self_attn.v_proj.weight": "model-00005-of-00007.safetensors",
115
- "llm2vec.layers.2.input_layernorm.weight": "model-00002-of-00007.safetensors",
116
- "llm2vec.layers.2.mlp.down_proj.weight": "model-00002-of-00007.safetensors",
117
- "llm2vec.layers.2.mlp.gate_proj.weight": "model-00002-of-00007.safetensors",
118
- "llm2vec.layers.2.mlp.up_proj.weight": "model-00002-of-00007.safetensors",
119
- "llm2vec.layers.2.post_attention_layernorm.weight": "model-00002-of-00007.safetensors",
120
- "llm2vec.layers.2.self_attn.k_proj.weight": "model-00002-of-00007.safetensors",
121
- "llm2vec.layers.2.self_attn.o_proj.weight": "model-00002-of-00007.safetensors",
122
- "llm2vec.layers.2.self_attn.q_proj.weight": "model-00002-of-00007.safetensors",
123
- "llm2vec.layers.2.self_attn.v_proj.weight": "model-00002-of-00007.safetensors",
124
- "llm2vec.layers.20.input_layernorm.weight": "model-00005-of-00007.safetensors",
125
- "llm2vec.layers.20.mlp.down_proj.weight": "model-00005-of-00007.safetensors",
126
- "llm2vec.layers.20.mlp.gate_proj.weight": "model-00005-of-00007.safetensors",
127
- "llm2vec.layers.20.mlp.up_proj.weight": "model-00005-of-00007.safetensors",
128
- "llm2vec.layers.20.post_attention_layernorm.weight": "model-00005-of-00007.safetensors",
129
- "llm2vec.layers.20.self_attn.k_proj.weight": "model-00005-of-00007.safetensors",
130
- "llm2vec.layers.20.self_attn.o_proj.weight": "model-00005-of-00007.safetensors",
131
- "llm2vec.layers.20.self_attn.q_proj.weight": "model-00005-of-00007.safetensors",
132
- "llm2vec.layers.20.self_attn.v_proj.weight": "model-00005-of-00007.safetensors",
133
- "llm2vec.layers.21.input_layernorm.weight": "model-00005-of-00007.safetensors",
134
- "llm2vec.layers.21.mlp.down_proj.weight": "model-00005-of-00007.safetensors",
135
- "llm2vec.layers.21.mlp.gate_proj.weight": "model-00005-of-00007.safetensors",
136
- "llm2vec.layers.21.mlp.up_proj.weight": "model-00005-of-00007.safetensors",
137
- "llm2vec.layers.21.post_attention_layernorm.weight": "model-00005-of-00007.safetensors",
138
- "llm2vec.layers.21.self_attn.k_proj.weight": "model-00005-of-00007.safetensors",
139
- "llm2vec.layers.21.self_attn.o_proj.weight": "model-00005-of-00007.safetensors",
140
- "llm2vec.layers.21.self_attn.q_proj.weight": "model-00005-of-00007.safetensors",
141
- "llm2vec.layers.21.self_attn.v_proj.weight": "model-00005-of-00007.safetensors",
142
- "llm2vec.layers.22.input_layernorm.weight": "model-00005-of-00007.safetensors",
143
- "llm2vec.layers.22.mlp.down_proj.weight": "model-00005-of-00007.safetensors",
144
- "llm2vec.layers.22.mlp.gate_proj.weight": "model-00005-of-00007.safetensors",
145
- "llm2vec.layers.22.mlp.up_proj.weight": "model-00005-of-00007.safetensors",
146
- "llm2vec.layers.22.post_attention_layernorm.weight": "model-00005-of-00007.safetensors",
147
- "llm2vec.layers.22.self_attn.k_proj.weight": "model-00005-of-00007.safetensors",
148
- "llm2vec.layers.22.self_attn.o_proj.weight": "model-00005-of-00007.safetensors",
149
- "llm2vec.layers.22.self_attn.q_proj.weight": "model-00005-of-00007.safetensors",
150
- "llm2vec.layers.22.self_attn.v_proj.weight": "model-00005-of-00007.safetensors",
151
- "llm2vec.layers.23.input_layernorm.weight": "model-00006-of-00007.safetensors",
152
- "llm2vec.layers.23.mlp.down_proj.weight": "model-00006-of-00007.safetensors",
153
- "llm2vec.layers.23.mlp.gate_proj.weight": "model-00005-of-00007.safetensors",
154
- "llm2vec.layers.23.mlp.up_proj.weight": "model-00005-of-00007.safetensors",
155
- "llm2vec.layers.23.post_attention_layernorm.weight": "model-00006-of-00007.safetensors",
156
- "llm2vec.layers.23.self_attn.k_proj.weight": "model-00005-of-00007.safetensors",
157
- "llm2vec.layers.23.self_attn.o_proj.weight": "model-00005-of-00007.safetensors",
158
- "llm2vec.layers.23.self_attn.q_proj.weight": "model-00005-of-00007.safetensors",
159
- "llm2vec.layers.23.self_attn.v_proj.weight": "model-00005-of-00007.safetensors",
160
- "llm2vec.layers.24.input_layernorm.weight": "model-00006-of-00007.safetensors",
161
- "llm2vec.layers.24.mlp.down_proj.weight": "model-00006-of-00007.safetensors",
162
- "llm2vec.layers.24.mlp.gate_proj.weight": "model-00006-of-00007.safetensors",
163
- "llm2vec.layers.24.mlp.up_proj.weight": "model-00006-of-00007.safetensors",
164
- "llm2vec.layers.24.post_attention_layernorm.weight": "model-00006-of-00007.safetensors",
165
- "llm2vec.layers.24.self_attn.k_proj.weight": "model-00006-of-00007.safetensors",
166
- "llm2vec.layers.24.self_attn.o_proj.weight": "model-00006-of-00007.safetensors",
167
- "llm2vec.layers.24.self_attn.q_proj.weight": "model-00006-of-00007.safetensors",
168
- "llm2vec.layers.24.self_attn.v_proj.weight": "model-00006-of-00007.safetensors",
169
- "llm2vec.layers.25.input_layernorm.weight": "model-00006-of-00007.safetensors",
170
- "llm2vec.layers.25.mlp.down_proj.weight": "model-00006-of-00007.safetensors",
171
- "llm2vec.layers.25.mlp.gate_proj.weight": "model-00006-of-00007.safetensors",
172
- "llm2vec.layers.25.mlp.up_proj.weight": "model-00006-of-00007.safetensors",
173
- "llm2vec.layers.25.post_attention_layernorm.weight": "model-00006-of-00007.safetensors",
174
- "llm2vec.layers.25.self_attn.k_proj.weight": "model-00006-of-00007.safetensors",
175
- "llm2vec.layers.25.self_attn.o_proj.weight": "model-00006-of-00007.safetensors",
176
- "llm2vec.layers.25.self_attn.q_proj.weight": "model-00006-of-00007.safetensors",
177
- "llm2vec.layers.25.self_attn.v_proj.weight": "model-00006-of-00007.safetensors",
178
- "llm2vec.layers.26.input_layernorm.weight": "model-00006-of-00007.safetensors",
179
- "llm2vec.layers.26.mlp.down_proj.weight": "model-00006-of-00007.safetensors",
180
- "llm2vec.layers.26.mlp.gate_proj.weight": "model-00006-of-00007.safetensors",
181
- "llm2vec.layers.26.mlp.up_proj.weight": "model-00006-of-00007.safetensors",
182
- "llm2vec.layers.26.post_attention_layernorm.weight": "model-00006-of-00007.safetensors",
183
- "llm2vec.layers.26.self_attn.k_proj.weight": "model-00006-of-00007.safetensors",
184
- "llm2vec.layers.26.self_attn.o_proj.weight": "model-00006-of-00007.safetensors",
185
- "llm2vec.layers.26.self_attn.q_proj.weight": "model-00006-of-00007.safetensors",
186
- "llm2vec.layers.26.self_attn.v_proj.weight": "model-00006-of-00007.safetensors",
187
- "llm2vec.layers.27.input_layernorm.weight": "model-00006-of-00007.safetensors",
188
- "llm2vec.layers.27.mlp.down_proj.weight": "model-00006-of-00007.safetensors",
189
- "llm2vec.layers.27.mlp.gate_proj.weight": "model-00006-of-00007.safetensors",
190
- "llm2vec.layers.27.mlp.up_proj.weight": "model-00006-of-00007.safetensors",
191
- "llm2vec.layers.27.post_attention_layernorm.weight": "model-00006-of-00007.safetensors",
192
- "llm2vec.layers.27.self_attn.k_proj.weight": "model-00006-of-00007.safetensors",
193
- "llm2vec.layers.27.self_attn.o_proj.weight": "model-00006-of-00007.safetensors",
194
- "llm2vec.layers.27.self_attn.q_proj.weight": "model-00006-of-00007.safetensors",
195
- "llm2vec.layers.27.self_attn.v_proj.weight": "model-00006-of-00007.safetensors",
196
- "llm2vec.layers.28.input_layernorm.weight": "model-00006-of-00007.safetensors",
197
- "llm2vec.layers.28.mlp.down_proj.weight": "model-00006-of-00007.safetensors",
198
- "llm2vec.layers.28.mlp.gate_proj.weight": "model-00006-of-00007.safetensors",
199
- "llm2vec.layers.28.mlp.up_proj.weight": "model-00006-of-00007.safetensors",
200
- "llm2vec.layers.28.post_attention_layernorm.weight": "model-00006-of-00007.safetensors",
201
- "llm2vec.layers.28.self_attn.k_proj.weight": "model-00006-of-00007.safetensors",
202
- "llm2vec.layers.28.self_attn.o_proj.weight": "model-00006-of-00007.safetensors",
203
- "llm2vec.layers.28.self_attn.q_proj.weight": "model-00006-of-00007.safetensors",
204
- "llm2vec.layers.28.self_attn.v_proj.weight": "model-00006-of-00007.safetensors",
205
- "llm2vec.layers.29.input_layernorm.weight": "model-00007-of-00007.safetensors",
206
- "llm2vec.layers.29.mlp.down_proj.weight": "model-00007-of-00007.safetensors",
207
- "llm2vec.layers.29.mlp.gate_proj.weight": "model-00006-of-00007.safetensors",
208
- "llm2vec.layers.29.mlp.up_proj.weight": "model-00007-of-00007.safetensors",
209
- "llm2vec.layers.29.post_attention_layernorm.weight": "model-00007-of-00007.safetensors",
210
- "llm2vec.layers.29.self_attn.k_proj.weight": "model-00006-of-00007.safetensors",
211
- "llm2vec.layers.29.self_attn.o_proj.weight": "model-00006-of-00007.safetensors",
212
- "llm2vec.layers.29.self_attn.q_proj.weight": "model-00006-of-00007.safetensors",
213
- "llm2vec.layers.29.self_attn.v_proj.weight": "model-00006-of-00007.safetensors",
214
- "llm2vec.layers.3.input_layernorm.weight": "model-00002-of-00007.safetensors",
215
- "llm2vec.layers.3.mlp.down_proj.weight": "model-00002-of-00007.safetensors",
216
- "llm2vec.layers.3.mlp.gate_proj.weight": "model-00002-of-00007.safetensors",
217
- "llm2vec.layers.3.mlp.up_proj.weight": "model-00002-of-00007.safetensors",
218
- "llm2vec.layers.3.post_attention_layernorm.weight": "model-00002-of-00007.safetensors",
219
- "llm2vec.layers.3.self_attn.k_proj.weight": "model-00002-of-00007.safetensors",
220
- "llm2vec.layers.3.self_attn.o_proj.weight": "model-00002-of-00007.safetensors",
221
- "llm2vec.layers.3.self_attn.q_proj.weight": "model-00002-of-00007.safetensors",
222
- "llm2vec.layers.3.self_attn.v_proj.weight": "model-00002-of-00007.safetensors",
223
- "llm2vec.layers.30.input_layernorm.weight": "model-00007-of-00007.safetensors",
224
- "llm2vec.layers.30.mlp.down_proj.weight": "model-00007-of-00007.safetensors",
225
- "llm2vec.layers.30.mlp.gate_proj.weight": "model-00007-of-00007.safetensors",
226
- "llm2vec.layers.30.mlp.up_proj.weight": "model-00007-of-00007.safetensors",
227
- "llm2vec.layers.30.post_attention_layernorm.weight": "model-00007-of-00007.safetensors",
228
- "llm2vec.layers.30.self_attn.k_proj.weight": "model-00007-of-00007.safetensors",
229
- "llm2vec.layers.30.self_attn.o_proj.weight": "model-00007-of-00007.safetensors",
230
- "llm2vec.layers.30.self_attn.q_proj.weight": "model-00007-of-00007.safetensors",
231
- "llm2vec.layers.30.self_attn.v_proj.weight": "model-00007-of-00007.safetensors",
232
- "llm2vec.layers.31.input_layernorm.weight": "model-00007-of-00007.safetensors",
233
- "llm2vec.layers.31.mlp.down_proj.weight": "model-00007-of-00007.safetensors",
234
- "llm2vec.layers.31.mlp.gate_proj.weight": "model-00007-of-00007.safetensors",
235
- "llm2vec.layers.31.mlp.up_proj.weight": "model-00007-of-00007.safetensors",
236
- "llm2vec.layers.31.post_attention_layernorm.weight": "model-00007-of-00007.safetensors",
237
- "llm2vec.layers.31.self_attn.k_proj.weight": "model-00007-of-00007.safetensors",
238
- "llm2vec.layers.31.self_attn.o_proj.weight": "model-00007-of-00007.safetensors",
239
- "llm2vec.layers.31.self_attn.q_proj.weight": "model-00007-of-00007.safetensors",
240
- "llm2vec.layers.31.self_attn.v_proj.weight": "model-00007-of-00007.safetensors",
241
- "llm2vec.layers.4.input_layernorm.weight": "model-00002-of-00007.safetensors",
242
- "llm2vec.layers.4.mlp.down_proj.weight": "model-00002-of-00007.safetensors",
243
- "llm2vec.layers.4.mlp.gate_proj.weight": "model-00002-of-00007.safetensors",
244
- "llm2vec.layers.4.mlp.up_proj.weight": "model-00002-of-00007.safetensors",
245
- "llm2vec.layers.4.post_attention_layernorm.weight": "model-00002-of-00007.safetensors",
246
- "llm2vec.layers.4.self_attn.k_proj.weight": "model-00002-of-00007.safetensors",
247
- "llm2vec.layers.4.self_attn.o_proj.weight": "model-00002-of-00007.safetensors",
248
- "llm2vec.layers.4.self_attn.q_proj.weight": "model-00002-of-00007.safetensors",
249
- "llm2vec.layers.4.self_attn.v_proj.weight": "model-00002-of-00007.safetensors",
250
- "llm2vec.layers.5.input_layernorm.weight": "model-00002-of-00007.safetensors",
251
- "llm2vec.layers.5.mlp.down_proj.weight": "model-00002-of-00007.safetensors",
252
- "llm2vec.layers.5.mlp.gate_proj.weight": "model-00002-of-00007.safetensors",
253
- "llm2vec.layers.5.mlp.up_proj.weight": "model-00002-of-00007.safetensors",
254
- "llm2vec.layers.5.post_attention_layernorm.weight": "model-00002-of-00007.safetensors",
255
- "llm2vec.layers.5.self_attn.k_proj.weight": "model-00002-of-00007.safetensors",
256
- "llm2vec.layers.5.self_attn.o_proj.weight": "model-00002-of-00007.safetensors",
257
- "llm2vec.layers.5.self_attn.q_proj.weight": "model-00002-of-00007.safetensors",
258
- "llm2vec.layers.5.self_attn.v_proj.weight": "model-00002-of-00007.safetensors",
259
- "llm2vec.layers.6.input_layernorm.weight": "model-00003-of-00007.safetensors",
260
- "llm2vec.layers.6.mlp.down_proj.weight": "model-00003-of-00007.safetensors",
261
- "llm2vec.layers.6.mlp.gate_proj.weight": "model-00002-of-00007.safetensors",
262
- "llm2vec.layers.6.mlp.up_proj.weight": "model-00002-of-00007.safetensors",
263
- "llm2vec.layers.6.post_attention_layernorm.weight": "model-00003-of-00007.safetensors",
264
- "llm2vec.layers.6.self_attn.k_proj.weight": "model-00002-of-00007.safetensors",
265
- "llm2vec.layers.6.self_attn.o_proj.weight": "model-00002-of-00007.safetensors",
266
- "llm2vec.layers.6.self_attn.q_proj.weight": "model-00002-of-00007.safetensors",
267
- "llm2vec.layers.6.self_attn.v_proj.weight": "model-00002-of-00007.safetensors",
268
- "llm2vec.layers.7.input_layernorm.weight": "model-00003-of-00007.safetensors",
269
- "llm2vec.layers.7.mlp.down_proj.weight": "model-00003-of-00007.safetensors",
270
- "llm2vec.layers.7.mlp.gate_proj.weight": "model-00003-of-00007.safetensors",
271
- "llm2vec.layers.7.mlp.up_proj.weight": "model-00003-of-00007.safetensors",
272
- "llm2vec.layers.7.post_attention_layernorm.weight": "model-00003-of-00007.safetensors",
273
- "llm2vec.layers.7.self_attn.k_proj.weight": "model-00003-of-00007.safetensors",
274
- "llm2vec.layers.7.self_attn.o_proj.weight": "model-00003-of-00007.safetensors",
275
- "llm2vec.layers.7.self_attn.q_proj.weight": "model-00003-of-00007.safetensors",
276
- "llm2vec.layers.7.self_attn.v_proj.weight": "model-00003-of-00007.safetensors",
277
- "llm2vec.layers.8.input_layernorm.weight": "model-00003-of-00007.safetensors",
278
- "llm2vec.layers.8.mlp.down_proj.weight": "model-00003-of-00007.safetensors",
279
- "llm2vec.layers.8.mlp.gate_proj.weight": "model-00003-of-00007.safetensors",
280
- "llm2vec.layers.8.mlp.up_proj.weight": "model-00003-of-00007.safetensors",
281
- "llm2vec.layers.8.post_attention_layernorm.weight": "model-00003-of-00007.safetensors",
282
- "llm2vec.layers.8.self_attn.k_proj.weight": "model-00003-of-00007.safetensors",
283
- "llm2vec.layers.8.self_attn.o_proj.weight": "model-00003-of-00007.safetensors",
284
- "llm2vec.layers.8.self_attn.q_proj.weight": "model-00003-of-00007.safetensors",
285
- "llm2vec.layers.8.self_attn.v_proj.weight": "model-00003-of-00007.safetensors",
286
- "llm2vec.layers.9.input_layernorm.weight": "model-00003-of-00007.safetensors",
287
- "llm2vec.layers.9.mlp.down_proj.weight": "model-00003-of-00007.safetensors",
288
- "llm2vec.layers.9.mlp.gate_proj.weight": "model-00003-of-00007.safetensors",
289
- "llm2vec.layers.9.mlp.up_proj.weight": "model-00003-of-00007.safetensors",
290
- "llm2vec.layers.9.post_attention_layernorm.weight": "model-00003-of-00007.safetensors",
291
- "llm2vec.layers.9.self_attn.k_proj.weight": "model-00003-of-00007.safetensors",
292
- "llm2vec.layers.9.self_attn.o_proj.weight": "model-00003-of-00007.safetensors",
293
- "llm2vec.layers.9.self_attn.q_proj.weight": "model-00003-of-00007.safetensors",
294
- "llm2vec.layers.9.self_attn.v_proj.weight": "model-00003-of-00007.safetensors",
295
- "llm2vec.norm.weight": "model-00007-of-00007.safetensors",
296
- "nllb_encoder.embed_tokens.weight": "model-00001-of-00007.safetensors",
297
- "nllb_encoder.layer_norm.bias": "model-00001-of-00007.safetensors",
298
- "nllb_encoder.layer_norm.weight": "model-00001-of-00007.safetensors",
299
- "nllb_encoder.layers.0.fc1.bias": "model-00001-of-00007.safetensors",
300
- "nllb_encoder.layers.0.fc1.weight": "model-00001-of-00007.safetensors",
301
- "nllb_encoder.layers.0.fc2.bias": "model-00001-of-00007.safetensors",
302
- "nllb_encoder.layers.0.fc2.weight": "model-00001-of-00007.safetensors",
303
- "nllb_encoder.layers.0.final_layer_norm.bias": "model-00001-of-00007.safetensors",
304
- "nllb_encoder.layers.0.final_layer_norm.weight": "model-00001-of-00007.safetensors",
305
- "nllb_encoder.layers.0.self_attn.k_proj.bias": "model-00001-of-00007.safetensors",
306
- "nllb_encoder.layers.0.self_attn.k_proj.weight": "model-00001-of-00007.safetensors",
307
- "nllb_encoder.layers.0.self_attn.out_proj.bias": "model-00001-of-00007.safetensors",
308
- "nllb_encoder.layers.0.self_attn.out_proj.weight": "model-00001-of-00007.safetensors",
309
- "nllb_encoder.layers.0.self_attn.q_proj.bias": "model-00001-of-00007.safetensors",
310
- "nllb_encoder.layers.0.self_attn.q_proj.weight": "model-00001-of-00007.safetensors",
311
- "nllb_encoder.layers.0.self_attn.v_proj.bias": "model-00001-of-00007.safetensors",
312
- "nllb_encoder.layers.0.self_attn.v_proj.weight": "model-00001-of-00007.safetensors",
313
- "nllb_encoder.layers.0.self_attn_layer_norm.bias": "model-00001-of-00007.safetensors",
314
- "nllb_encoder.layers.0.self_attn_layer_norm.weight": "model-00001-of-00007.safetensors",
315
- "nllb_encoder.layers.1.fc1.bias": "model-00001-of-00007.safetensors",
316
- "nllb_encoder.layers.1.fc1.weight": "model-00001-of-00007.safetensors",
317
- "nllb_encoder.layers.1.fc2.bias": "model-00001-of-00007.safetensors",
318
- "nllb_encoder.layers.1.fc2.weight": "model-00001-of-00007.safetensors",
319
- "nllb_encoder.layers.1.final_layer_norm.bias": "model-00001-of-00007.safetensors",
320
- "nllb_encoder.layers.1.final_layer_norm.weight": "model-00001-of-00007.safetensors",
321
- "nllb_encoder.layers.1.self_attn.k_proj.bias": "model-00001-of-00007.safetensors",
322
- "nllb_encoder.layers.1.self_attn.k_proj.weight": "model-00001-of-00007.safetensors",
323
- "nllb_encoder.layers.1.self_attn.out_proj.bias": "model-00001-of-00007.safetensors",
324
- "nllb_encoder.layers.1.self_attn.out_proj.weight": "model-00001-of-00007.safetensors",
325
- "nllb_encoder.layers.1.self_attn.q_proj.bias": "model-00001-of-00007.safetensors",
326
- "nllb_encoder.layers.1.self_attn.q_proj.weight": "model-00001-of-00007.safetensors",
327
- "nllb_encoder.layers.1.self_attn.v_proj.bias": "model-00001-of-00007.safetensors",
328
- "nllb_encoder.layers.1.self_attn.v_proj.weight": "model-00001-of-00007.safetensors",
329
- "nllb_encoder.layers.1.self_attn_layer_norm.bias": "model-00001-of-00007.safetensors",
330
- "nllb_encoder.layers.1.self_attn_layer_norm.weight": "model-00001-of-00007.safetensors",
331
- "nllb_encoder.layers.10.fc1.bias": "model-00001-of-00007.safetensors",
332
- "nllb_encoder.layers.10.fc1.weight": "model-00001-of-00007.safetensors",
333
- "nllb_encoder.layers.10.fc2.bias": "model-00001-of-00007.safetensors",
334
- "nllb_encoder.layers.10.fc2.weight": "model-00001-of-00007.safetensors",
335
- "nllb_encoder.layers.10.final_layer_norm.bias": "model-00001-of-00007.safetensors",
336
- "nllb_encoder.layers.10.final_layer_norm.weight": "model-00001-of-00007.safetensors",
337
- "nllb_encoder.layers.10.self_attn.k_proj.bias": "model-00001-of-00007.safetensors",
338
- "nllb_encoder.layers.10.self_attn.k_proj.weight": "model-00001-of-00007.safetensors",
339
- "nllb_encoder.layers.10.self_attn.out_proj.bias": "model-00001-of-00007.safetensors",
340
- "nllb_encoder.layers.10.self_attn.out_proj.weight": "model-00001-of-00007.safetensors",
341
- "nllb_encoder.layers.10.self_attn.q_proj.bias": "model-00001-of-00007.safetensors",
342
- "nllb_encoder.layers.10.self_attn.q_proj.weight": "model-00001-of-00007.safetensors",
343
- "nllb_encoder.layers.10.self_attn.v_proj.bias": "model-00001-of-00007.safetensors",
344
- "nllb_encoder.layers.10.self_attn.v_proj.weight": "model-00001-of-00007.safetensors",
345
- "nllb_encoder.layers.10.self_attn_layer_norm.bias": "model-00001-of-00007.safetensors",
346
- "nllb_encoder.layers.10.self_attn_layer_norm.weight": "model-00001-of-00007.safetensors",
347
- "nllb_encoder.layers.11.fc1.bias": "model-00001-of-00007.safetensors",
348
- "nllb_encoder.layers.11.fc1.weight": "model-00001-of-00007.safetensors",
349
- "nllb_encoder.layers.11.fc2.bias": "model-00001-of-00007.safetensors",
350
- "nllb_encoder.layers.11.fc2.weight": "model-00001-of-00007.safetensors",
351
- "nllb_encoder.layers.11.final_layer_norm.bias": "model-00001-of-00007.safetensors",
352
- "nllb_encoder.layers.11.final_layer_norm.weight": "model-00001-of-00007.safetensors",
353
- "nllb_encoder.layers.11.self_attn.k_proj.bias": "model-00001-of-00007.safetensors",
354
- "nllb_encoder.layers.11.self_attn.k_proj.weight": "model-00001-of-00007.safetensors",
355
- "nllb_encoder.layers.11.self_attn.out_proj.bias": "model-00001-of-00007.safetensors",
356
- "nllb_encoder.layers.11.self_attn.out_proj.weight": "model-00001-of-00007.safetensors",
357
- "nllb_encoder.layers.11.self_attn.q_proj.bias": "model-00001-of-00007.safetensors",
358
- "nllb_encoder.layers.11.self_attn.q_proj.weight": "model-00001-of-00007.safetensors",
359
- "nllb_encoder.layers.11.self_attn.v_proj.bias": "model-00001-of-00007.safetensors",
360
- "nllb_encoder.layers.11.self_attn.v_proj.weight": "model-00001-of-00007.safetensors",
361
- "nllb_encoder.layers.11.self_attn_layer_norm.bias": "model-00001-of-00007.safetensors",
362
- "nllb_encoder.layers.11.self_attn_layer_norm.weight": "model-00001-of-00007.safetensors",
363
- "nllb_encoder.layers.2.fc1.bias": "model-00001-of-00007.safetensors",
364
- "nllb_encoder.layers.2.fc1.weight": "model-00001-of-00007.safetensors",
365
- "nllb_encoder.layers.2.fc2.bias": "model-00001-of-00007.safetensors",
366
- "nllb_encoder.layers.2.fc2.weight": "model-00001-of-00007.safetensors",
367
- "nllb_encoder.layers.2.final_layer_norm.bias": "model-00001-of-00007.safetensors",
368
- "nllb_encoder.layers.2.final_layer_norm.weight": "model-00001-of-00007.safetensors",
369
- "nllb_encoder.layers.2.self_attn.k_proj.bias": "model-00001-of-00007.safetensors",
370
- "nllb_encoder.layers.2.self_attn.k_proj.weight": "model-00001-of-00007.safetensors",
371
- "nllb_encoder.layers.2.self_attn.out_proj.bias": "model-00001-of-00007.safetensors",
372
- "nllb_encoder.layers.2.self_attn.out_proj.weight": "model-00001-of-00007.safetensors",
373
- "nllb_encoder.layers.2.self_attn.q_proj.bias": "model-00001-of-00007.safetensors",
374
- "nllb_encoder.layers.2.self_attn.q_proj.weight": "model-00001-of-00007.safetensors",
375
- "nllb_encoder.layers.2.self_attn.v_proj.bias": "model-00001-of-00007.safetensors",
376
- "nllb_encoder.layers.2.self_attn.v_proj.weight": "model-00001-of-00007.safetensors",
377
- "nllb_encoder.layers.2.self_attn_layer_norm.bias": "model-00001-of-00007.safetensors",
378
- "nllb_encoder.layers.2.self_attn_layer_norm.weight": "model-00001-of-00007.safetensors",
379
- "nllb_encoder.layers.3.fc1.bias": "model-00001-of-00007.safetensors",
380
- "nllb_encoder.layers.3.fc1.weight": "model-00001-of-00007.safetensors",
381
- "nllb_encoder.layers.3.fc2.bias": "model-00001-of-00007.safetensors",
382
- "nllb_encoder.layers.3.fc2.weight": "model-00001-of-00007.safetensors",
383
- "nllb_encoder.layers.3.final_layer_norm.bias": "model-00001-of-00007.safetensors",
384
- "nllb_encoder.layers.3.final_layer_norm.weight": "model-00001-of-00007.safetensors",
385
- "nllb_encoder.layers.3.self_attn.k_proj.bias": "model-00001-of-00007.safetensors",
386
- "nllb_encoder.layers.3.self_attn.k_proj.weight": "model-00001-of-00007.safetensors",
387
- "nllb_encoder.layers.3.self_attn.out_proj.bias": "model-00001-of-00007.safetensors",
388
- "nllb_encoder.layers.3.self_attn.out_proj.weight": "model-00001-of-00007.safetensors",
389
- "nllb_encoder.layers.3.self_attn.q_proj.bias": "model-00001-of-00007.safetensors",
390
- "nllb_encoder.layers.3.self_attn.q_proj.weight": "model-00001-of-00007.safetensors",
391
- "nllb_encoder.layers.3.self_attn.v_proj.bias": "model-00001-of-00007.safetensors",
392
- "nllb_encoder.layers.3.self_attn.v_proj.weight": "model-00001-of-00007.safetensors",
393
- "nllb_encoder.layers.3.self_attn_layer_norm.bias": "model-00001-of-00007.safetensors",
394
- "nllb_encoder.layers.3.self_attn_layer_norm.weight": "model-00001-of-00007.safetensors",
395
- "nllb_encoder.layers.4.fc1.bias": "model-00001-of-00007.safetensors",
396
- "nllb_encoder.layers.4.fc1.weight": "model-00001-of-00007.safetensors",
397
- "nllb_encoder.layers.4.fc2.bias": "model-00001-of-00007.safetensors",
398
- "nllb_encoder.layers.4.fc2.weight": "model-00001-of-00007.safetensors",
399
- "nllb_encoder.layers.4.final_layer_norm.bias": "model-00001-of-00007.safetensors",
400
- "nllb_encoder.layers.4.final_layer_norm.weight": "model-00001-of-00007.safetensors",
401
- "nllb_encoder.layers.4.self_attn.k_proj.bias": "model-00001-of-00007.safetensors",
402
- "nllb_encoder.layers.4.self_attn.k_proj.weight": "model-00001-of-00007.safetensors",
403
- "nllb_encoder.layers.4.self_attn.out_proj.bias": "model-00001-of-00007.safetensors",
404
- "nllb_encoder.layers.4.self_attn.out_proj.weight": "model-00001-of-00007.safetensors",
405
- "nllb_encoder.layers.4.self_attn.q_proj.bias": "model-00001-of-00007.safetensors",
406
- "nllb_encoder.layers.4.self_attn.q_proj.weight": "model-00001-of-00007.safetensors",
407
- "nllb_encoder.layers.4.self_attn.v_proj.bias": "model-00001-of-00007.safetensors",
408
- "nllb_encoder.layers.4.self_attn.v_proj.weight": "model-00001-of-00007.safetensors",
409
- "nllb_encoder.layers.4.self_attn_layer_norm.bias": "model-00001-of-00007.safetensors",
410
- "nllb_encoder.layers.4.self_attn_layer_norm.weight": "model-00001-of-00007.safetensors",
411
- "nllb_encoder.layers.5.fc1.bias": "model-00001-of-00007.safetensors",
412
- "nllb_encoder.layers.5.fc1.weight": "model-00001-of-00007.safetensors",
413
- "nllb_encoder.layers.5.fc2.bias": "model-00001-of-00007.safetensors",
414
- "nllb_encoder.layers.5.fc2.weight": "model-00001-of-00007.safetensors",
415
- "nllb_encoder.layers.5.final_layer_norm.bias": "model-00001-of-00007.safetensors",
416
- "nllb_encoder.layers.5.final_layer_norm.weight": "model-00001-of-00007.safetensors",
417
- "nllb_encoder.layers.5.self_attn.k_proj.bias": "model-00001-of-00007.safetensors",
418
- "nllb_encoder.layers.5.self_attn.k_proj.weight": "model-00001-of-00007.safetensors",
419
- "nllb_encoder.layers.5.self_attn.out_proj.bias": "model-00001-of-00007.safetensors",
420
- "nllb_encoder.layers.5.self_attn.out_proj.weight": "model-00001-of-00007.safetensors",
421
- "nllb_encoder.layers.5.self_attn.q_proj.bias": "model-00001-of-00007.safetensors",
422
- "nllb_encoder.layers.5.self_attn.q_proj.weight": "model-00001-of-00007.safetensors",
423
- "nllb_encoder.layers.5.self_attn.v_proj.bias": "model-00001-of-00007.safetensors",
424
- "nllb_encoder.layers.5.self_attn.v_proj.weight": "model-00001-of-00007.safetensors",
425
- "nllb_encoder.layers.5.self_attn_layer_norm.bias": "model-00001-of-00007.safetensors",
426
- "nllb_encoder.layers.5.self_attn_layer_norm.weight": "model-00001-of-00007.safetensors",
427
- "nllb_encoder.layers.6.fc1.bias": "model-00001-of-00007.safetensors",
428
- "nllb_encoder.layers.6.fc1.weight": "model-00001-of-00007.safetensors",
429
- "nllb_encoder.layers.6.fc2.bias": "model-00001-of-00007.safetensors",
430
- "nllb_encoder.layers.6.fc2.weight": "model-00001-of-00007.safetensors",
431
- "nllb_encoder.layers.6.final_layer_norm.bias": "model-00001-of-00007.safetensors",
432
- "nllb_encoder.layers.6.final_layer_norm.weight": "model-00001-of-00007.safetensors",
433
- "nllb_encoder.layers.6.self_attn.k_proj.bias": "model-00001-of-00007.safetensors",
434
- "nllb_encoder.layers.6.self_attn.k_proj.weight": "model-00001-of-00007.safetensors",
435
- "nllb_encoder.layers.6.self_attn.out_proj.bias": "model-00001-of-00007.safetensors",
436
- "nllb_encoder.layers.6.self_attn.out_proj.weight": "model-00001-of-00007.safetensors",
437
- "nllb_encoder.layers.6.self_attn.q_proj.bias": "model-00001-of-00007.safetensors",
438
- "nllb_encoder.layers.6.self_attn.q_proj.weight": "model-00001-of-00007.safetensors",
439
- "nllb_encoder.layers.6.self_attn.v_proj.bias": "model-00001-of-00007.safetensors",
440
- "nllb_encoder.layers.6.self_attn.v_proj.weight": "model-00001-of-00007.safetensors",
441
- "nllb_encoder.layers.6.self_attn_layer_norm.bias": "model-00001-of-00007.safetensors",
442
- "nllb_encoder.layers.6.self_attn_layer_norm.weight": "model-00001-of-00007.safetensors",
443
- "nllb_encoder.layers.7.fc1.bias": "model-00001-of-00007.safetensors",
444
- "nllb_encoder.layers.7.fc1.weight": "model-00001-of-00007.safetensors",
445
- "nllb_encoder.layers.7.fc2.bias": "model-00001-of-00007.safetensors",
446
- "nllb_encoder.layers.7.fc2.weight": "model-00001-of-00007.safetensors",
447
- "nllb_encoder.layers.7.final_layer_norm.bias": "model-00001-of-00007.safetensors",
448
- "nllb_encoder.layers.7.final_layer_norm.weight": "model-00001-of-00007.safetensors",
449
- "nllb_encoder.layers.7.self_attn.k_proj.bias": "model-00001-of-00007.safetensors",
450
- "nllb_encoder.layers.7.self_attn.k_proj.weight": "model-00001-of-00007.safetensors",
451
- "nllb_encoder.layers.7.self_attn.out_proj.bias": "model-00001-of-00007.safetensors",
452
- "nllb_encoder.layers.7.self_attn.out_proj.weight": "model-00001-of-00007.safetensors",
453
- "nllb_encoder.layers.7.self_attn.q_proj.bias": "model-00001-of-00007.safetensors",
454
- "nllb_encoder.layers.7.self_attn.q_proj.weight": "model-00001-of-00007.safetensors",
455
- "nllb_encoder.layers.7.self_attn.v_proj.bias": "model-00001-of-00007.safetensors",
456
- "nllb_encoder.layers.7.self_attn.v_proj.weight": "model-00001-of-00007.safetensors",
457
- "nllb_encoder.layers.7.self_attn_layer_norm.bias": "model-00001-of-00007.safetensors",
458
- "nllb_encoder.layers.7.self_attn_layer_norm.weight": "model-00001-of-00007.safetensors",
459
- "nllb_encoder.layers.8.fc1.bias": "model-00001-of-00007.safetensors",
460
- "nllb_encoder.layers.8.fc1.weight": "model-00001-of-00007.safetensors",
461
- "nllb_encoder.layers.8.fc2.bias": "model-00001-of-00007.safetensors",
462
- "nllb_encoder.layers.8.fc2.weight": "model-00001-of-00007.safetensors",
463
- "nllb_encoder.layers.8.final_layer_norm.bias": "model-00001-of-00007.safetensors",
464
- "nllb_encoder.layers.8.final_layer_norm.weight": "model-00001-of-00007.safetensors",
465
- "nllb_encoder.layers.8.self_attn.k_proj.bias": "model-00001-of-00007.safetensors",
466
- "nllb_encoder.layers.8.self_attn.k_proj.weight": "model-00001-of-00007.safetensors",
467
- "nllb_encoder.layers.8.self_attn.out_proj.bias": "model-00001-of-00007.safetensors",
468
- "nllb_encoder.layers.8.self_attn.out_proj.weight": "model-00001-of-00007.safetensors",
469
- "nllb_encoder.layers.8.self_attn.q_proj.bias": "model-00001-of-00007.safetensors",
470
- "nllb_encoder.layers.8.self_attn.q_proj.weight": "model-00001-of-00007.safetensors",
471
- "nllb_encoder.layers.8.self_attn.v_proj.bias": "model-00001-of-00007.safetensors",
472
- "nllb_encoder.layers.8.self_attn.v_proj.weight": "model-00001-of-00007.safetensors",
473
- "nllb_encoder.layers.8.self_attn_layer_norm.bias": "model-00001-of-00007.safetensors",
474
- "nllb_encoder.layers.8.self_attn_layer_norm.weight": "model-00001-of-00007.safetensors",
475
- "nllb_encoder.layers.9.fc1.bias": "model-00001-of-00007.safetensors",
476
- "nllb_encoder.layers.9.fc1.weight": "model-00001-of-00007.safetensors",
477
- "nllb_encoder.layers.9.fc2.bias": "model-00001-of-00007.safetensors",
478
- "nllb_encoder.layers.9.fc2.weight": "model-00001-of-00007.safetensors",
479
- "nllb_encoder.layers.9.final_layer_norm.bias": "model-00001-of-00007.safetensors",
480
- "nllb_encoder.layers.9.final_layer_norm.weight": "model-00001-of-00007.safetensors",
481
- "nllb_encoder.layers.9.self_attn.k_proj.bias": "model-00001-of-00007.safetensors",
482
- "nllb_encoder.layers.9.self_attn.k_proj.weight": "model-00001-of-00007.safetensors",
483
- "nllb_encoder.layers.9.self_attn.out_proj.bias": "model-00001-of-00007.safetensors",
484
- "nllb_encoder.layers.9.self_attn.out_proj.weight": "model-00001-of-00007.safetensors",
485
- "nllb_encoder.layers.9.self_attn.q_proj.bias": "model-00001-of-00007.safetensors",
486
- "nllb_encoder.layers.9.self_attn.q_proj.weight": "model-00001-of-00007.safetensors",
487
- "nllb_encoder.layers.9.self_attn.v_proj.bias": "model-00001-of-00007.safetensors",
488
- "nllb_encoder.layers.9.self_attn.v_proj.weight": "model-00001-of-00007.safetensors",
489
- "nllb_encoder.layers.9.self_attn_layer_norm.bias": "model-00001-of-00007.safetensors",
490
- "nllb_encoder.layers.9.self_attn_layer_norm.weight": "model-00001-of-00007.safetensors",
491
- "up_proj.weight": "model-00007-of-00007.safetensors"
492
  }
493
  }
 
1
  {
2
  "metadata": {
3
+ "total_size": 15845261312
4
  },
5
  "weight_map": {
6
+ "llm2vec.embed_tokens.weight": "model-00001-of-00004.safetensors",
7
+ "llm2vec.layers.0.input_layernorm.weight": "model-00001-of-00004.safetensors",
8
+ "llm2vec.layers.0.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
9
+ "llm2vec.layers.0.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
10
+ "llm2vec.layers.0.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
11
+ "llm2vec.layers.0.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
12
+ "llm2vec.layers.0.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
13
+ "llm2vec.layers.0.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
14
+ "llm2vec.layers.0.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
15
+ "llm2vec.layers.0.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
16
+ "llm2vec.layers.1.input_layernorm.weight": "model-00001-of-00004.safetensors",
17
+ "llm2vec.layers.1.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
18
+ "llm2vec.layers.1.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
19
+ "llm2vec.layers.1.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
20
+ "llm2vec.layers.1.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
21
+ "llm2vec.layers.1.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
22
+ "llm2vec.layers.1.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
23
+ "llm2vec.layers.1.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
24
+ "llm2vec.layers.1.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
25
+ "llm2vec.layers.10.input_layernorm.weight": "model-00002-of-00004.safetensors",
26
+ "llm2vec.layers.10.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
27
+ "llm2vec.layers.10.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
28
+ "llm2vec.layers.10.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
29
+ "llm2vec.layers.10.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
30
+ "llm2vec.layers.10.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
31
+ "llm2vec.layers.10.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
32
+ "llm2vec.layers.10.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
33
+ "llm2vec.layers.10.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
34
+ "llm2vec.layers.11.input_layernorm.weight": "model-00002-of-00004.safetensors",
35
+ "llm2vec.layers.11.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
36
+ "llm2vec.layers.11.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
37
+ "llm2vec.layers.11.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
38
+ "llm2vec.layers.11.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
39
+ "llm2vec.layers.11.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
40
+ "llm2vec.layers.11.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
41
+ "llm2vec.layers.11.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
42
+ "llm2vec.layers.11.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
43
+ "llm2vec.layers.12.input_layernorm.weight": "model-00002-of-00004.safetensors",
44
+ "llm2vec.layers.12.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
45
+ "llm2vec.layers.12.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
46
+ "llm2vec.layers.12.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
47
+ "llm2vec.layers.12.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
48
+ "llm2vec.layers.12.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
49
+ "llm2vec.layers.12.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
50
+ "llm2vec.layers.12.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
51
+ "llm2vec.layers.12.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
52
+ "llm2vec.layers.13.input_layernorm.weight": "model-00002-of-00004.safetensors",
53
+ "llm2vec.layers.13.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
54
+ "llm2vec.layers.13.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
55
+ "llm2vec.layers.13.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
56
+ "llm2vec.layers.13.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
57
+ "llm2vec.layers.13.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
58
+ "llm2vec.layers.13.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
59
+ "llm2vec.layers.13.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
60
+ "llm2vec.layers.13.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
61
+ "llm2vec.layers.14.input_layernorm.weight": "model-00002-of-00004.safetensors",
62
+ "llm2vec.layers.14.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
63
+ "llm2vec.layers.14.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
64
+ "llm2vec.layers.14.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
65
+ "llm2vec.layers.14.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
66
+ "llm2vec.layers.14.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
67
+ "llm2vec.layers.14.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
68
+ "llm2vec.layers.14.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
69
+ "llm2vec.layers.14.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
70
+ "llm2vec.layers.15.input_layernorm.weight": "model-00002-of-00004.safetensors",
71
+ "llm2vec.layers.15.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
72
+ "llm2vec.layers.15.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
73
+ "llm2vec.layers.15.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
74
+ "llm2vec.layers.15.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
75
+ "llm2vec.layers.15.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
76
+ "llm2vec.layers.15.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
77
+ "llm2vec.layers.15.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
78
+ "llm2vec.layers.15.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
79
+ "llm2vec.layers.16.input_layernorm.weight": "model-00002-of-00004.safetensors",
80
+ "llm2vec.layers.16.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
81
+ "llm2vec.layers.16.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
82
+ "llm2vec.layers.16.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
83
+ "llm2vec.layers.16.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
84
+ "llm2vec.layers.16.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
85
+ "llm2vec.layers.16.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
86
+ "llm2vec.layers.16.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
87
+ "llm2vec.layers.16.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
88
+ "llm2vec.layers.17.input_layernorm.weight": "model-00002-of-00004.safetensors",
89
+ "llm2vec.layers.17.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
90
+ "llm2vec.layers.17.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
91
+ "llm2vec.layers.17.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
92
+ "llm2vec.layers.17.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
93
+ "llm2vec.layers.17.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
94
+ "llm2vec.layers.17.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
95
+ "llm2vec.layers.17.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
96
+ "llm2vec.layers.17.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
97
+ "llm2vec.layers.18.input_layernorm.weight": "model-00003-of-00004.safetensors",
98
+ "llm2vec.layers.18.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
99
+ "llm2vec.layers.18.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
100
+ "llm2vec.layers.18.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
101
+ "llm2vec.layers.18.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
102
+ "llm2vec.layers.18.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
103
+ "llm2vec.layers.18.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
104
+ "llm2vec.layers.18.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
105
+ "llm2vec.layers.18.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
106
+ "llm2vec.layers.19.input_layernorm.weight": "model-00003-of-00004.safetensors",
107
+ "llm2vec.layers.19.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
108
+ "llm2vec.layers.19.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
109
+ "llm2vec.layers.19.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
110
+ "llm2vec.layers.19.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
111
+ "llm2vec.layers.19.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
112
+ "llm2vec.layers.19.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
113
+ "llm2vec.layers.19.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
114
+ "llm2vec.layers.19.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
115
+ "llm2vec.layers.2.input_layernorm.weight": "model-00001-of-00004.safetensors",
116
+ "llm2vec.layers.2.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
117
+ "llm2vec.layers.2.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
118
+ "llm2vec.layers.2.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
119
+ "llm2vec.layers.2.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
120
+ "llm2vec.layers.2.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
121
+ "llm2vec.layers.2.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
122
+ "llm2vec.layers.2.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
123
+ "llm2vec.layers.2.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
124
+ "llm2vec.layers.20.input_layernorm.weight": "model-00003-of-00004.safetensors",
125
+ "llm2vec.layers.20.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
126
+ "llm2vec.layers.20.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
127
+ "llm2vec.layers.20.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
128
+ "llm2vec.layers.20.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
129
+ "llm2vec.layers.20.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
130
+ "llm2vec.layers.20.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
131
+ "llm2vec.layers.20.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
132
+ "llm2vec.layers.20.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
133
+ "llm2vec.layers.21.input_layernorm.weight": "model-00003-of-00004.safetensors",
134
+ "llm2vec.layers.21.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
135
+ "llm2vec.layers.21.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
136
+ "llm2vec.layers.21.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
137
+ "llm2vec.layers.21.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
138
+ "llm2vec.layers.21.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
139
+ "llm2vec.layers.21.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
140
+ "llm2vec.layers.21.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
141
+ "llm2vec.layers.21.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
142
+ "llm2vec.layers.22.input_layernorm.weight": "model-00003-of-00004.safetensors",
143
+ "llm2vec.layers.22.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
144
+ "llm2vec.layers.22.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
145
+ "llm2vec.layers.22.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
146
+ "llm2vec.layers.22.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
147
+ "llm2vec.layers.22.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
148
+ "llm2vec.layers.22.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
149
+ "llm2vec.layers.22.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
150
+ "llm2vec.layers.22.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
151
+ "llm2vec.layers.23.input_layernorm.weight": "model-00003-of-00004.safetensors",
152
+ "llm2vec.layers.23.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
153
+ "llm2vec.layers.23.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
154
+ "llm2vec.layers.23.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
155
+ "llm2vec.layers.23.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
156
+ "llm2vec.layers.23.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
157
+ "llm2vec.layers.23.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
158
+ "llm2vec.layers.23.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
159
+ "llm2vec.layers.23.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
160
+ "llm2vec.layers.24.input_layernorm.weight": "model-00003-of-00004.safetensors",
161
+ "llm2vec.layers.24.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
162
+ "llm2vec.layers.24.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
163
+ "llm2vec.layers.24.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
164
+ "llm2vec.layers.24.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
165
+ "llm2vec.layers.24.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
166
+ "llm2vec.layers.24.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
167
+ "llm2vec.layers.24.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
168
+ "llm2vec.layers.24.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
169
+ "llm2vec.layers.25.input_layernorm.weight": "model-00003-of-00004.safetensors",
170
+ "llm2vec.layers.25.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
171
+ "llm2vec.layers.25.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
172
+ "llm2vec.layers.25.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
173
+ "llm2vec.layers.25.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
174
+ "llm2vec.layers.25.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
175
+ "llm2vec.layers.25.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
176
+ "llm2vec.layers.25.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
177
+ "llm2vec.layers.25.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
178
+ "llm2vec.layers.26.input_layernorm.weight": "model-00003-of-00004.safetensors",
179
+ "llm2vec.layers.26.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
180
+ "llm2vec.layers.26.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
181
+ "llm2vec.layers.26.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
182
+ "llm2vec.layers.26.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
183
+ "llm2vec.layers.26.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
184
+ "llm2vec.layers.26.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
185
+ "llm2vec.layers.26.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
186
+ "llm2vec.layers.26.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
187
+ "llm2vec.layers.27.input_layernorm.weight": "model-00003-of-00004.safetensors",
188
+ "llm2vec.layers.27.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
189
+ "llm2vec.layers.27.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
190
+ "llm2vec.layers.27.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
191
+ "llm2vec.layers.27.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
192
+ "llm2vec.layers.27.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
193
+ "llm2vec.layers.27.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
194
+ "llm2vec.layers.27.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
195
+ "llm2vec.layers.27.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
196
+ "llm2vec.layers.28.input_layernorm.weight": "model-00003-of-00004.safetensors",
197
+ "llm2vec.layers.28.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
198
+ "llm2vec.layers.28.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
199
+ "llm2vec.layers.28.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
200
+ "llm2vec.layers.28.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
201
+ "llm2vec.layers.28.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
202
+ "llm2vec.layers.28.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
203
+ "llm2vec.layers.28.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
204
+ "llm2vec.layers.28.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
205
+ "llm2vec.layers.29.input_layernorm.weight": "model-00004-of-00004.safetensors",
206
+ "llm2vec.layers.29.mlp.down_proj.weight": "model-00004-of-00004.safetensors",
207
+ "llm2vec.layers.29.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
208
+ "llm2vec.layers.29.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
209
+ "llm2vec.layers.29.post_attention_layernorm.weight": "model-00004-of-00004.safetensors",
210
+ "llm2vec.layers.29.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
211
+ "llm2vec.layers.29.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
212
+ "llm2vec.layers.29.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
213
+ "llm2vec.layers.29.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
214
+ "llm2vec.layers.3.input_layernorm.weight": "model-00001-of-00004.safetensors",
215
+ "llm2vec.layers.3.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
216
+ "llm2vec.layers.3.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
217
+ "llm2vec.layers.3.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
218
+ "llm2vec.layers.3.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
219
+ "llm2vec.layers.3.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
220
+ "llm2vec.layers.3.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
221
+ "llm2vec.layers.3.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
222
+ "llm2vec.layers.3.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
223
+ "llm2vec.layers.30.input_layernorm.weight": "model-00004-of-00004.safetensors",
224
+ "llm2vec.layers.30.mlp.down_proj.weight": "model-00004-of-00004.safetensors",
225
+ "llm2vec.layers.30.mlp.gate_proj.weight": "model-00004-of-00004.safetensors",
226
+ "llm2vec.layers.30.mlp.up_proj.weight": "model-00004-of-00004.safetensors",
227
+ "llm2vec.layers.30.post_attention_layernorm.weight": "model-00004-of-00004.safetensors",
228
+ "llm2vec.layers.30.self_attn.k_proj.weight": "model-00004-of-00004.safetensors",
229
+ "llm2vec.layers.30.self_attn.o_proj.weight": "model-00004-of-00004.safetensors",
230
+ "llm2vec.layers.30.self_attn.q_proj.weight": "model-00004-of-00004.safetensors",
231
+ "llm2vec.layers.30.self_attn.v_proj.weight": "model-00004-of-00004.safetensors",
232
+ "llm2vec.layers.31.input_layernorm.weight": "model-00004-of-00004.safetensors",
233
+ "llm2vec.layers.31.mlp.down_proj.weight": "model-00004-of-00004.safetensors",
234
+ "llm2vec.layers.31.mlp.gate_proj.weight": "model-00004-of-00004.safetensors",
235
+ "llm2vec.layers.31.mlp.up_proj.weight": "model-00004-of-00004.safetensors",
236
+ "llm2vec.layers.31.post_attention_layernorm.weight": "model-00004-of-00004.safetensors",
237
+ "llm2vec.layers.31.self_attn.k_proj.weight": "model-00004-of-00004.safetensors",
238
+ "llm2vec.layers.31.self_attn.o_proj.weight": "model-00004-of-00004.safetensors",
239
+ "llm2vec.layers.31.self_attn.q_proj.weight": "model-00004-of-00004.safetensors",
240
+ "llm2vec.layers.31.self_attn.v_proj.weight": "model-00004-of-00004.safetensors",
241
+ "llm2vec.layers.4.input_layernorm.weight": "model-00001-of-00004.safetensors",
242
+ "llm2vec.layers.4.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
243
+ "llm2vec.layers.4.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
244
+ "llm2vec.layers.4.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
245
+ "llm2vec.layers.4.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
246
+ "llm2vec.layers.4.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
247
+ "llm2vec.layers.4.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
248
+ "llm2vec.layers.4.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
249
+ "llm2vec.layers.4.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
250
+ "llm2vec.layers.5.input_layernorm.weight": "model-00001-of-00004.safetensors",
251
+ "llm2vec.layers.5.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
252
+ "llm2vec.layers.5.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
253
+ "llm2vec.layers.5.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
254
+ "llm2vec.layers.5.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
255
+ "llm2vec.layers.5.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
256
+ "llm2vec.layers.5.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
257
+ "llm2vec.layers.5.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
258
+ "llm2vec.layers.5.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
259
+ "llm2vec.layers.6.input_layernorm.weight": "model-00001-of-00004.safetensors",
260
+ "llm2vec.layers.6.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
261
+ "llm2vec.layers.6.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
262
+ "llm2vec.layers.6.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
263
+ "llm2vec.layers.6.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
264
+ "llm2vec.layers.6.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
265
+ "llm2vec.layers.6.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
266
+ "llm2vec.layers.6.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
267
+ "llm2vec.layers.6.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
268
+ "llm2vec.layers.7.input_layernorm.weight": "model-00002-of-00004.safetensors",
269
+ "llm2vec.layers.7.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
270
+ "llm2vec.layers.7.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
271
+ "llm2vec.layers.7.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
272
+ "llm2vec.layers.7.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
273
+ "llm2vec.layers.7.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
274
+ "llm2vec.layers.7.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
275
+ "llm2vec.layers.7.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
276
+ "llm2vec.layers.7.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
277
+ "llm2vec.layers.8.input_layernorm.weight": "model-00002-of-00004.safetensors",
278
+ "llm2vec.layers.8.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
279
+ "llm2vec.layers.8.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
280
+ "llm2vec.layers.8.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
281
+ "llm2vec.layers.8.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
282
+ "llm2vec.layers.8.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
283
+ "llm2vec.layers.8.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
284
+ "llm2vec.layers.8.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
285
+ "llm2vec.layers.8.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
286
+ "llm2vec.layers.9.input_layernorm.weight": "model-00002-of-00004.safetensors",
287
+ "llm2vec.layers.9.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
288
+ "llm2vec.layers.9.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
289
+ "llm2vec.layers.9.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
290
+ "llm2vec.layers.9.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
291
+ "llm2vec.layers.9.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
292
+ "llm2vec.layers.9.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
293
+ "llm2vec.layers.9.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
294
+ "llm2vec.layers.9.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
295
+ "llm2vec.norm.weight": "model-00004-of-00004.safetensors",
296
+ "nllb_encoder.embed_tokens.weight": "model-00001-of-00004.safetensors",
297
+ "nllb_encoder.layer_norm.bias": "model-00001-of-00004.safetensors",
298
+ "nllb_encoder.layer_norm.weight": "model-00001-of-00004.safetensors",
299
+ "nllb_encoder.layers.0.fc1.bias": "model-00001-of-00004.safetensors",
300
+ "nllb_encoder.layers.0.fc1.weight": "model-00001-of-00004.safetensors",
301
+ "nllb_encoder.layers.0.fc2.bias": "model-00001-of-00004.safetensors",
302
+ "nllb_encoder.layers.0.fc2.weight": "model-00001-of-00004.safetensors",
303
+ "nllb_encoder.layers.0.final_layer_norm.bias": "model-00001-of-00004.safetensors",
304
+ "nllb_encoder.layers.0.final_layer_norm.weight": "model-00001-of-00004.safetensors",
305
+ "nllb_encoder.layers.0.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
306
+ "nllb_encoder.layers.0.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
307
+ "nllb_encoder.layers.0.self_attn.out_proj.bias": "model-00001-of-00004.safetensors",
308
+ "nllb_encoder.layers.0.self_attn.out_proj.weight": "model-00001-of-00004.safetensors",
309
+ "nllb_encoder.layers.0.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
310
+ "nllb_encoder.layers.0.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
311
+ "nllb_encoder.layers.0.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
312
+ "nllb_encoder.layers.0.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
313
+ "nllb_encoder.layers.0.self_attn_layer_norm.bias": "model-00001-of-00004.safetensors",
314
+ "nllb_encoder.layers.0.self_attn_layer_norm.weight": "model-00001-of-00004.safetensors",
315
+ "nllb_encoder.layers.1.fc1.bias": "model-00001-of-00004.safetensors",
316
+ "nllb_encoder.layers.1.fc1.weight": "model-00001-of-00004.safetensors",
317
+ "nllb_encoder.layers.1.fc2.bias": "model-00001-of-00004.safetensors",
318
+ "nllb_encoder.layers.1.fc2.weight": "model-00001-of-00004.safetensors",
319
+ "nllb_encoder.layers.1.final_layer_norm.bias": "model-00001-of-00004.safetensors",
320
+ "nllb_encoder.layers.1.final_layer_norm.weight": "model-00001-of-00004.safetensors",
321
+ "nllb_encoder.layers.1.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
322
+ "nllb_encoder.layers.1.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
323
+ "nllb_encoder.layers.1.self_attn.out_proj.bias": "model-00001-of-00004.safetensors",
324
+ "nllb_encoder.layers.1.self_attn.out_proj.weight": "model-00001-of-00004.safetensors",
325
+ "nllb_encoder.layers.1.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
326
+ "nllb_encoder.layers.1.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
327
+ "nllb_encoder.layers.1.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
328
+ "nllb_encoder.layers.1.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
329
+ "nllb_encoder.layers.1.self_attn_layer_norm.bias": "model-00001-of-00004.safetensors",
330
+ "nllb_encoder.layers.1.self_attn_layer_norm.weight": "model-00001-of-00004.safetensors",
331
+ "nllb_encoder.layers.10.fc1.bias": "model-00001-of-00004.safetensors",
332
+ "nllb_encoder.layers.10.fc1.weight": "model-00001-of-00004.safetensors",
333
+ "nllb_encoder.layers.10.fc2.bias": "model-00001-of-00004.safetensors",
334
+ "nllb_encoder.layers.10.fc2.weight": "model-00001-of-00004.safetensors",
335
+ "nllb_encoder.layers.10.final_layer_norm.bias": "model-00001-of-00004.safetensors",
336
+ "nllb_encoder.layers.10.final_layer_norm.weight": "model-00001-of-00004.safetensors",
337
+ "nllb_encoder.layers.10.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
338
+ "nllb_encoder.layers.10.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
339
+ "nllb_encoder.layers.10.self_attn.out_proj.bias": "model-00001-of-00004.safetensors",
340
+ "nllb_encoder.layers.10.self_attn.out_proj.weight": "model-00001-of-00004.safetensors",
341
+ "nllb_encoder.layers.10.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
342
+ "nllb_encoder.layers.10.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
343
+ "nllb_encoder.layers.10.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
344
+ "nllb_encoder.layers.10.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
345
+ "nllb_encoder.layers.10.self_attn_layer_norm.bias": "model-00001-of-00004.safetensors",
346
+ "nllb_encoder.layers.10.self_attn_layer_norm.weight": "model-00001-of-00004.safetensors",
347
+ "nllb_encoder.layers.11.fc1.bias": "model-00001-of-00004.safetensors",
348
+ "nllb_encoder.layers.11.fc1.weight": "model-00001-of-00004.safetensors",
349
+ "nllb_encoder.layers.11.fc2.bias": "model-00001-of-00004.safetensors",
350
+ "nllb_encoder.layers.11.fc2.weight": "model-00001-of-00004.safetensors",
351
+ "nllb_encoder.layers.11.final_layer_norm.bias": "model-00001-of-00004.safetensors",
352
+ "nllb_encoder.layers.11.final_layer_norm.weight": "model-00001-of-00004.safetensors",
353
+ "nllb_encoder.layers.11.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
354
+ "nllb_encoder.layers.11.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
355
+ "nllb_encoder.layers.11.self_attn.out_proj.bias": "model-00001-of-00004.safetensors",
356
+ "nllb_encoder.layers.11.self_attn.out_proj.weight": "model-00001-of-00004.safetensors",
357
+ "nllb_encoder.layers.11.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
358
+ "nllb_encoder.layers.11.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
359
+ "nllb_encoder.layers.11.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
360
+ "nllb_encoder.layers.11.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
361
+ "nllb_encoder.layers.11.self_attn_layer_norm.bias": "model-00001-of-00004.safetensors",
362
+ "nllb_encoder.layers.11.self_attn_layer_norm.weight": "model-00001-of-00004.safetensors",
363
+ "nllb_encoder.layers.2.fc1.bias": "model-00001-of-00004.safetensors",
364
+ "nllb_encoder.layers.2.fc1.weight": "model-00001-of-00004.safetensors",
365
+ "nllb_encoder.layers.2.fc2.bias": "model-00001-of-00004.safetensors",
366
+ "nllb_encoder.layers.2.fc2.weight": "model-00001-of-00004.safetensors",
367
+ "nllb_encoder.layers.2.final_layer_norm.bias": "model-00001-of-00004.safetensors",
368
+ "nllb_encoder.layers.2.final_layer_norm.weight": "model-00001-of-00004.safetensors",
369
+ "nllb_encoder.layers.2.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
370
+ "nllb_encoder.layers.2.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
371
+ "nllb_encoder.layers.2.self_attn.out_proj.bias": "model-00001-of-00004.safetensors",
372
+ "nllb_encoder.layers.2.self_attn.out_proj.weight": "model-00001-of-00004.safetensors",
373
+ "nllb_encoder.layers.2.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
374
+ "nllb_encoder.layers.2.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
375
+ "nllb_encoder.layers.2.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
376
+ "nllb_encoder.layers.2.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
377
+ "nllb_encoder.layers.2.self_attn_layer_norm.bias": "model-00001-of-00004.safetensors",
378
+ "nllb_encoder.layers.2.self_attn_layer_norm.weight": "model-00001-of-00004.safetensors",
379
+ "nllb_encoder.layers.3.fc1.bias": "model-00001-of-00004.safetensors",
380
+ "nllb_encoder.layers.3.fc1.weight": "model-00001-of-00004.safetensors",
381
+ "nllb_encoder.layers.3.fc2.bias": "model-00001-of-00004.safetensors",
382
+ "nllb_encoder.layers.3.fc2.weight": "model-00001-of-00004.safetensors",
383
+ "nllb_encoder.layers.3.final_layer_norm.bias": "model-00001-of-00004.safetensors",
384
+ "nllb_encoder.layers.3.final_layer_norm.weight": "model-00001-of-00004.safetensors",
385
+ "nllb_encoder.layers.3.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
386
+ "nllb_encoder.layers.3.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
387
+ "nllb_encoder.layers.3.self_attn.out_proj.bias": "model-00001-of-00004.safetensors",
388
+ "nllb_encoder.layers.3.self_attn.out_proj.weight": "model-00001-of-00004.safetensors",
389
+ "nllb_encoder.layers.3.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
390
+ "nllb_encoder.layers.3.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
391
+ "nllb_encoder.layers.3.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
392
+ "nllb_encoder.layers.3.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
393
+ "nllb_encoder.layers.3.self_attn_layer_norm.bias": "model-00001-of-00004.safetensors",
394
+ "nllb_encoder.layers.3.self_attn_layer_norm.weight": "model-00001-of-00004.safetensors",
395
+ "nllb_encoder.layers.4.fc1.bias": "model-00001-of-00004.safetensors",
396
+ "nllb_encoder.layers.4.fc1.weight": "model-00001-of-00004.safetensors",
397
+ "nllb_encoder.layers.4.fc2.bias": "model-00001-of-00004.safetensors",
398
+ "nllb_encoder.layers.4.fc2.weight": "model-00001-of-00004.safetensors",
399
+ "nllb_encoder.layers.4.final_layer_norm.bias": "model-00001-of-00004.safetensors",
400
+ "nllb_encoder.layers.4.final_layer_norm.weight": "model-00001-of-00004.safetensors",
401
+ "nllb_encoder.layers.4.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
402
+ "nllb_encoder.layers.4.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
403
+ "nllb_encoder.layers.4.self_attn.out_proj.bias": "model-00001-of-00004.safetensors",
404
+ "nllb_encoder.layers.4.self_attn.out_proj.weight": "model-00001-of-00004.safetensors",
405
+ "nllb_encoder.layers.4.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
406
+ "nllb_encoder.layers.4.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
407
+ "nllb_encoder.layers.4.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
408
+ "nllb_encoder.layers.4.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
409
+ "nllb_encoder.layers.4.self_attn_layer_norm.bias": "model-00001-of-00004.safetensors",
410
+ "nllb_encoder.layers.4.self_attn_layer_norm.weight": "model-00001-of-00004.safetensors",
411
+ "nllb_encoder.layers.5.fc1.bias": "model-00001-of-00004.safetensors",
412
+ "nllb_encoder.layers.5.fc1.weight": "model-00001-of-00004.safetensors",
413
+ "nllb_encoder.layers.5.fc2.bias": "model-00001-of-00004.safetensors",
414
+ "nllb_encoder.layers.5.fc2.weight": "model-00001-of-00004.safetensors",
415
+ "nllb_encoder.layers.5.final_layer_norm.bias": "model-00001-of-00004.safetensors",
416
+ "nllb_encoder.layers.5.final_layer_norm.weight": "model-00001-of-00004.safetensors",
417
+ "nllb_encoder.layers.5.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
418
+ "nllb_encoder.layers.5.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
419
+ "nllb_encoder.layers.5.self_attn.out_proj.bias": "model-00001-of-00004.safetensors",
420
+ "nllb_encoder.layers.5.self_attn.out_proj.weight": "model-00001-of-00004.safetensors",
421
+ "nllb_encoder.layers.5.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
422
+ "nllb_encoder.layers.5.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
423
+ "nllb_encoder.layers.5.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
424
+ "nllb_encoder.layers.5.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
425
+ "nllb_encoder.layers.5.self_attn_layer_norm.bias": "model-00001-of-00004.safetensors",
426
+ "nllb_encoder.layers.5.self_attn_layer_norm.weight": "model-00001-of-00004.safetensors",
427
+ "nllb_encoder.layers.6.fc1.bias": "model-00001-of-00004.safetensors",
428
+ "nllb_encoder.layers.6.fc1.weight": "model-00001-of-00004.safetensors",
429
+ "nllb_encoder.layers.6.fc2.bias": "model-00001-of-00004.safetensors",
430
+ "nllb_encoder.layers.6.fc2.weight": "model-00001-of-00004.safetensors",
431
+ "nllb_encoder.layers.6.final_layer_norm.bias": "model-00001-of-00004.safetensors",
432
+ "nllb_encoder.layers.6.final_layer_norm.weight": "model-00001-of-00004.safetensors",
433
+ "nllb_encoder.layers.6.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
434
+ "nllb_encoder.layers.6.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
435
+ "nllb_encoder.layers.6.self_attn.out_proj.bias": "model-00001-of-00004.safetensors",
436
+ "nllb_encoder.layers.6.self_attn.out_proj.weight": "model-00001-of-00004.safetensors",
437
+ "nllb_encoder.layers.6.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
438
+ "nllb_encoder.layers.6.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
439
+ "nllb_encoder.layers.6.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
440
+ "nllb_encoder.layers.6.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
441
+ "nllb_encoder.layers.6.self_attn_layer_norm.bias": "model-00001-of-00004.safetensors",
442
+ "nllb_encoder.layers.6.self_attn_layer_norm.weight": "model-00001-of-00004.safetensors",
443
+ "nllb_encoder.layers.7.fc1.bias": "model-00001-of-00004.safetensors",
444
+ "nllb_encoder.layers.7.fc1.weight": "model-00001-of-00004.safetensors",
445
+ "nllb_encoder.layers.7.fc2.bias": "model-00001-of-00004.safetensors",
446
+ "nllb_encoder.layers.7.fc2.weight": "model-00001-of-00004.safetensors",
447
+ "nllb_encoder.layers.7.final_layer_norm.bias": "model-00001-of-00004.safetensors",
448
+ "nllb_encoder.layers.7.final_layer_norm.weight": "model-00001-of-00004.safetensors",
449
+ "nllb_encoder.layers.7.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
450
+ "nllb_encoder.layers.7.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
451
+ "nllb_encoder.layers.7.self_attn.out_proj.bias": "model-00001-of-00004.safetensors",
452
+ "nllb_encoder.layers.7.self_attn.out_proj.weight": "model-00001-of-00004.safetensors",
453
+ "nllb_encoder.layers.7.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
454
+ "nllb_encoder.layers.7.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
455
+ "nllb_encoder.layers.7.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
456
+ "nllb_encoder.layers.7.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
457
+ "nllb_encoder.layers.7.self_attn_layer_norm.bias": "model-00001-of-00004.safetensors",
458
+ "nllb_encoder.layers.7.self_attn_layer_norm.weight": "model-00001-of-00004.safetensors",
459
+ "nllb_encoder.layers.8.fc1.bias": "model-00001-of-00004.safetensors",
460
+ "nllb_encoder.layers.8.fc1.weight": "model-00001-of-00004.safetensors",
461
+ "nllb_encoder.layers.8.fc2.bias": "model-00001-of-00004.safetensors",
462
+ "nllb_encoder.layers.8.fc2.weight": "model-00001-of-00004.safetensors",
463
+ "nllb_encoder.layers.8.final_layer_norm.bias": "model-00001-of-00004.safetensors",
464
+ "nllb_encoder.layers.8.final_layer_norm.weight": "model-00001-of-00004.safetensors",
465
+ "nllb_encoder.layers.8.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
466
+ "nllb_encoder.layers.8.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
467
+ "nllb_encoder.layers.8.self_attn.out_proj.bias": "model-00001-of-00004.safetensors",
468
+ "nllb_encoder.layers.8.self_attn.out_proj.weight": "model-00001-of-00004.safetensors",
469
+ "nllb_encoder.layers.8.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
470
+ "nllb_encoder.layers.8.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
471
+ "nllb_encoder.layers.8.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
472
+ "nllb_encoder.layers.8.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
473
+ "nllb_encoder.layers.8.self_attn_layer_norm.bias": "model-00001-of-00004.safetensors",
474
+ "nllb_encoder.layers.8.self_attn_layer_norm.weight": "model-00001-of-00004.safetensors",
475
+ "nllb_encoder.layers.9.fc1.bias": "model-00001-of-00004.safetensors",
476
+ "nllb_encoder.layers.9.fc1.weight": "model-00001-of-00004.safetensors",
477
+ "nllb_encoder.layers.9.fc2.bias": "model-00001-of-00004.safetensors",
478
+ "nllb_encoder.layers.9.fc2.weight": "model-00001-of-00004.safetensors",
479
+ "nllb_encoder.layers.9.final_layer_norm.bias": "model-00001-of-00004.safetensors",
480
+ "nllb_encoder.layers.9.final_layer_norm.weight": "model-00001-of-00004.safetensors",
481
+ "nllb_encoder.layers.9.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
482
+ "nllb_encoder.layers.9.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
483
+ "nllb_encoder.layers.9.self_attn.out_proj.bias": "model-00001-of-00004.safetensors",
484
+ "nllb_encoder.layers.9.self_attn.out_proj.weight": "model-00001-of-00004.safetensors",
485
+ "nllb_encoder.layers.9.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
486
+ "nllb_encoder.layers.9.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
487
+ "nllb_encoder.layers.9.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
488
+ "nllb_encoder.layers.9.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
489
+ "nllb_encoder.layers.9.self_attn_layer_norm.bias": "model-00001-of-00004.safetensors",
490
+ "nllb_encoder.layers.9.self_attn_layer_norm.weight": "model-00001-of-00004.safetensors",
491
+ "up_proj.weight": "model-00004-of-00004.safetensors"
492
  }
493
  }