Hugging Face's logo

wpferrell
/
phi-3.5-mini-instruct-bigsmall

phi3
bigsmall
compressed
lossless
custom_code
Model card Files
xet
 Community
1
wpferrell commited on
Commit
d534bca
·
verified ·
1 Parent(s): 94f36fd

Upload bigsmall.index.json with huggingface_hub

Browse files
Files changed (1) hide show
  1. bigsmall.index.json +215 -0
bigsmall.index.json ADDED
@@ -0,0 +1,215 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "metadata": {
3
+ "bigsmall_version": "2.0.1",
4
+ "container_version": 1,
5
+ "format": "bf16",
6
+ "mode": "balanced",
7
+ "model_type": "llm",
8
+ "total_size": 5017892507,
9
+ "total_raw_size": 7642159104,
10
+ "ratio_pct": 65.66066524803931,
11
+ "shard_count": 2,
12
+ "tensor_count": 195,
13
+ "shards": [
14
+ "model-00001-of-00002.bs",
15
+ "model-00002-of-00002.bs"
16
+ ]
17
+ },
18
+ "weight_map": {
19
+ "model.embed_tokens.weight": "model-00001-of-00002.bs",
20
+ "model.layers.0.input_layernorm.weight": "model-00001-of-00002.bs",
21
+ "model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.bs",
22
+ "model.layers.0.mlp.gate_up_proj.weight": "model-00001-of-00002.bs",
23
+ "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.bs",
24
+ "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.bs",
25
+ "model.layers.0.self_attn.qkv_proj.weight": "model-00001-of-00002.bs",
26
+ "model.layers.1.input_layernorm.weight": "model-00001-of-00002.bs",
27
+ "model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.bs",
28
+ "model.layers.1.mlp.gate_up_proj.weight": "model-00001-of-00002.bs",
29
+ "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.bs",
30
+ "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.bs",
31
+ "model.layers.1.self_attn.qkv_proj.weight": "model-00001-of-00002.bs",
32
+ "model.layers.10.input_layernorm.weight": "model-00001-of-00002.bs",
33
+ "model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.bs",
34
+ "model.layers.10.mlp.gate_up_proj.weight": "model-00001-of-00002.bs",
35
+ "model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.bs",
36
+ "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.bs",
37
+ "model.layers.10.self_attn.qkv_proj.weight": "model-00001-of-00002.bs",
38
+ "model.layers.11.input_layernorm.weight": "model-00001-of-00002.bs",
39
+ "model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.bs",
40
+ "model.layers.11.mlp.gate_up_proj.weight": "model-00001-of-00002.bs",
41
+ "model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.bs",
42
+ "model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.bs",
43
+ "model.layers.11.self_attn.qkv_proj.weight": "model-00001-of-00002.bs",
44
+ "model.layers.12.input_layernorm.weight": "model-00001-of-00002.bs",
45
+ "model.layers.12.mlp.down_proj.weight": "model-00001-of-00002.bs",
46
+ "model.layers.12.mlp.gate_up_proj.weight": "model-00001-of-00002.bs",
47
+ "model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.bs",
48
+ "model.layers.12.self_attn.o_proj.weight": "model-00001-of-00002.bs",
49
+ "model.layers.12.self_attn.qkv_proj.weight": "model-00001-of-00002.bs",
50
+ "model.layers.13.input_layernorm.weight": "model-00001-of-00002.bs",
51
+ "model.layers.13.mlp.down_proj.weight": "model-00001-of-00002.bs",
52
+ "model.layers.13.mlp.gate_up_proj.weight": "model-00001-of-00002.bs",
53
+ "model.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.bs",
54
+ "model.layers.13.self_attn.o_proj.weight": "model-00001-of-00002.bs",
55
+ "model.layers.13.self_attn.qkv_proj.weight": "model-00001-of-00002.bs",
56
+ "model.layers.14.input_layernorm.weight": "model-00001-of-00002.bs",
57
+ "model.layers.14.mlp.down_proj.weight": "model-00001-of-00002.bs",
58
+ "model.layers.14.mlp.gate_up_proj.weight": "model-00001-of-00002.bs",
59
+ "model.layers.14.post_attention_layernorm.weight": "model-00001-of-00002.bs",
60
+ "model.layers.14.self_attn.o_proj.weight": "model-00001-of-00002.bs",
61
+ "model.layers.14.self_attn.qkv_proj.weight": "model-00001-of-00002.bs",
62
+ "model.layers.15.input_layernorm.weight": "model-00001-of-00002.bs",
63
+ "model.layers.15.mlp.down_proj.weight": "model-00001-of-00002.bs",
64
+ "model.layers.15.mlp.gate_up_proj.weight": "model-00001-of-00002.bs",
65
+ "model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.bs",
66
+ "model.layers.15.self_attn.o_proj.weight": "model-00001-of-00002.bs",
67
+ "model.layers.15.self_attn.qkv_proj.weight": "model-00001-of-00002.bs",
68
+ "model.layers.16.input_layernorm.weight": "model-00001-of-00002.bs",
69
+ "model.layers.16.mlp.down_proj.weight": "model-00001-of-00002.bs",
70
+ "model.layers.16.mlp.gate_up_proj.weight": "model-00001-of-00002.bs",
71
+ "model.layers.16.post_attention_layernorm.weight": "model-00001-of-00002.bs",
72
+ "model.layers.16.self_attn.o_proj.weight": "model-00001-of-00002.bs",
73
+ "model.layers.16.self_attn.qkv_proj.weight": "model-00001-of-00002.bs",
74
+ "model.layers.17.input_layernorm.weight": "model-00001-of-00002.bs",
75
+ "model.layers.17.mlp.down_proj.weight": "model-00001-of-00002.bs",
76
+ "model.layers.17.mlp.gate_up_proj.weight": "model-00001-of-00002.bs",
77
+ "model.layers.17.post_attention_layernorm.weight": "model-00001-of-00002.bs",
78
+ "model.layers.17.self_attn.o_proj.weight": "model-00001-of-00002.bs",
79
+ "model.layers.17.self_attn.qkv_proj.weight": "model-00001-of-00002.bs",
80
+ "model.layers.18.input_layernorm.weight": "model-00001-of-00002.bs",
81
+ "model.layers.18.mlp.down_proj.weight": "model-00001-of-00002.bs",
82
+ "model.layers.18.mlp.gate_up_proj.weight": "model-00001-of-00002.bs",
83
+ "model.layers.18.post_attention_layernorm.weight": "model-00001-of-00002.bs",
84
+ "model.layers.18.self_attn.o_proj.weight": "model-00001-of-00002.bs",
85
+ "model.layers.18.self_attn.qkv_proj.weight": "model-00001-of-00002.bs",
86
+ "model.layers.19.input_layernorm.weight": "model-00001-of-00002.bs",
87
+ "model.layers.19.mlp.down_proj.weight": "model-00001-of-00002.bs",
88
+ "model.layers.19.mlp.gate_up_proj.weight": "model-00001-of-00002.bs",
89
+ "model.layers.19.post_attention_layernorm.weight": "model-00001-of-00002.bs",
90
+ "model.layers.19.self_attn.o_proj.weight": "model-00001-of-00002.bs",
91
+ "model.layers.19.self_attn.qkv_proj.weight": "model-00001-of-00002.bs",
92
+ "model.layers.2.input_layernorm.weight": "model-00001-of-00002.bs",
93
+ "model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.bs",
94
+ "model.layers.2.mlp.gate_up_proj.weight": "model-00001-of-00002.bs",
95
+ "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.bs",
96
+ "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.bs",
97
+ "model.layers.2.self_attn.qkv_proj.weight": "model-00001-of-00002.bs",
98
+ "model.layers.20.input_layernorm.weight": "model-00001-of-00002.bs",
99
+ "model.layers.20.mlp.down_proj.weight": "model-00001-of-00002.bs",
100
+ "model.layers.20.mlp.gate_up_proj.weight": "model-00001-of-00002.bs",
101
+ "model.layers.20.post_attention_layernorm.weight": "model-00001-of-00002.bs",
102
+ "model.layers.20.self_attn.o_proj.weight": "model-00001-of-00002.bs",
103
+ "model.layers.20.self_attn.qkv_proj.weight": "model-00001-of-00002.bs",
104
+ "model.layers.21.self_attn.o_proj.weight": "model-00001-of-00002.bs",
105
+ "model.layers.3.input_layernorm.weight": "model-00001-of-00002.bs",
106
+ "model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.bs",
107
+ "model.layers.3.mlp.gate_up_proj.weight": "model-00001-of-00002.bs",
108
+ "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.bs",
109
+ "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.bs",
110
+ "model.layers.3.self_attn.qkv_proj.weight": "model-00001-of-00002.bs",
111
+ "model.layers.4.input_layernorm.weight": "model-00001-of-00002.bs",
112
+ "model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.bs",
113
+ "model.layers.4.mlp.gate_up_proj.weight": "model-00001-of-00002.bs",
114
+ "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.bs",
115
+ "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.bs",
116
+ "model.layers.4.self_attn.qkv_proj.weight": "model-00001-of-00002.bs",
117
+ "model.layers.5.input_layernorm.weight": "model-00001-of-00002.bs",
118
+ "model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.bs",
119
+ "model.layers.5.mlp.gate_up_proj.weight": "model-00001-of-00002.bs",
120
+ "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.bs",
121
+ "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.bs",
122
+ "model.layers.5.self_attn.qkv_proj.weight": "model-00001-of-00002.bs",
123
+ "model.layers.6.input_layernorm.weight": "model-00001-of-00002.bs",
124
+ "model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.bs",
125
+ "model.layers.6.mlp.gate_up_proj.weight": "model-00001-of-00002.bs",
126
+ "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.bs",
127
+ "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.bs",
128
+ "model.layers.6.self_attn.qkv_proj.weight": "model-00001-of-00002.bs",
129
+ "model.layers.7.input_layernorm.weight": "model-00001-of-00002.bs",
130
+ "model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.bs",
131
+ "model.layers.7.mlp.gate_up_proj.weight": "model-00001-of-00002.bs",
132
+ "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.bs",
133
+ "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.bs",
134
+ "model.layers.7.self_attn.qkv_proj.weight": "model-00001-of-00002.bs",
135
+ "model.layers.8.input_layernorm.weight": "model-00001-of-00002.bs",
136
+ "model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.bs",
137
+ "model.layers.8.mlp.gate_up_proj.weight": "model-00001-of-00002.bs",
138
+ "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.bs",
139
+ "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.bs",
140
+ "model.layers.8.self_attn.qkv_proj.weight": "model-00001-of-00002.bs",
141
+ "model.layers.9.input_layernorm.weight": "model-00001-of-00002.bs",
142
+ "model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.bs",
143
+ "model.layers.9.mlp.gate_up_proj.weight": "model-00001-of-00002.bs",
144
+ "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.bs",
145
+ "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.bs",
146
+ "model.layers.9.self_attn.qkv_proj.weight": "model-00001-of-00002.bs",
147
+ "lm_head.weight": "model-00002-of-00002.bs",
148
+ "model.layers.21.input_layernorm.weight": "model-00002-of-00002.bs",
149
+ "model.layers.21.mlp.down_proj.weight": "model-00002-of-00002.bs",
150
+ "model.layers.21.mlp.gate_up_proj.weight": "model-00002-of-00002.bs",
151
+ "model.layers.21.post_attention_layernorm.weight": "model-00002-of-00002.bs",
152
+ "model.layers.21.self_attn.qkv_proj.weight": "model-00002-of-00002.bs",
153
+ "model.layers.22.input_layernorm.weight": "model-00002-of-00002.bs",
154
+ "model.layers.22.mlp.down_proj.weight": "model-00002-of-00002.bs",
155
+ "model.layers.22.mlp.gate_up_proj.weight": "model-00002-of-00002.bs",
156
+ "model.layers.22.post_attention_layernorm.weight": "model-00002-of-00002.bs",
157
+ "model.layers.22.self_attn.o_proj.weight": "model-00002-of-00002.bs",
158
+ "model.layers.22.self_attn.qkv_proj.weight": "model-00002-of-00002.bs",
159
+ "model.layers.23.input_layernorm.weight": "model-00002-of-00002.bs",
160
+ "model.layers.23.mlp.down_proj.weight": "model-00002-of-00002.bs",
161
+ "model.layers.23.mlp.gate_up_proj.weight": "model-00002-of-00002.bs",
162
+ "model.layers.23.post_attention_layernorm.weight": "model-00002-of-00002.bs",
163
+ "model.layers.23.self_attn.o_proj.weight": "model-00002-of-00002.bs",
164
+ "model.layers.23.self_attn.qkv_proj.weight": "model-00002-of-00002.bs",
165
+ "model.layers.24.input_layernorm.weight": "model-00002-of-00002.bs",
166
+ "model.layers.24.mlp.down_proj.weight": "model-00002-of-00002.bs",
167
+ "model.layers.24.mlp.gate_up_proj.weight": "model-00002-of-00002.bs",
168
+ "model.layers.24.post_attention_layernorm.weight": "model-00002-of-00002.bs",
169
+ "model.layers.24.self_attn.o_proj.weight": "model-00002-of-00002.bs",
170
+ "model.layers.24.self_attn.qkv_proj.weight": "model-00002-of-00002.bs",
171
+ "model.layers.25.input_layernorm.weight": "model-00002-of-00002.bs",
172
+ "model.layers.25.mlp.down_proj.weight": "model-00002-of-00002.bs",
173
+ "model.layers.25.mlp.gate_up_proj.weight": "model-00002-of-00002.bs",
174
+ "model.layers.25.post_attention_layernorm.weight": "model-00002-of-00002.bs",
175
+ "model.layers.25.self_attn.o_proj.weight": "model-00002-of-00002.bs",
176
+ "model.layers.25.self_attn.qkv_proj.weight": "model-00002-of-00002.bs",
177
+ "model.layers.26.input_layernorm.weight": "model-00002-of-00002.bs",
178
+ "model.layers.26.mlp.down_proj.weight": "model-00002-of-00002.bs",
179
+ "model.layers.26.mlp.gate_up_proj.weight": "model-00002-of-00002.bs",
180
+ "model.layers.26.post_attention_layernorm.weight": "model-00002-of-00002.bs",
181
+ "model.layers.26.self_attn.o_proj.weight": "model-00002-of-00002.bs",
182
+ "model.layers.26.self_attn.qkv_proj.weight": "model-00002-of-00002.bs",
183
+ "model.layers.27.input_layernorm.weight": "model-00002-of-00002.bs",
184
+ "model.layers.27.mlp.down_proj.weight": "model-00002-of-00002.bs",
185
+ "model.layers.27.mlp.gate_up_proj.weight": "model-00002-of-00002.bs",
186
+ "model.layers.27.post_attention_layernorm.weight": "model-00002-of-00002.bs",
187
+ "model.layers.27.self_attn.o_proj.weight": "model-00002-of-00002.bs",
188
+ "model.layers.27.self_attn.qkv_proj.weight": "model-00002-of-00002.bs",
189
+ "model.layers.28.input_layernorm.weight": "model-00002-of-00002.bs",
190
+ "model.layers.28.mlp.down_proj.weight": "model-00002-of-00002.bs",
191
+ "model.layers.28.mlp.gate_up_proj.weight": "model-00002-of-00002.bs",
192
+ "model.layers.28.post_attention_layernorm.weight": "model-00002-of-00002.bs",
193
+ "model.layers.28.self_attn.o_proj.weight": "model-00002-of-00002.bs",
194
+ "model.layers.28.self_attn.qkv_proj.weight": "model-00002-of-00002.bs",
195
+ "model.layers.29.input_layernorm.weight": "model-00002-of-00002.bs",
196
+ "model.layers.29.mlp.down_proj.weight": "model-00002-of-00002.bs",
197
+ "model.layers.29.mlp.gate_up_proj.weight": "model-00002-of-00002.bs",
198
+ "model.layers.29.post_attention_layernorm.weight": "model-00002-of-00002.bs",
199
+ "model.layers.29.self_attn.o_proj.weight": "model-00002-of-00002.bs",
200
+ "model.layers.29.self_attn.qkv_proj.weight": "model-00002-of-00002.bs",
201
+ "model.layers.30.input_layernorm.weight": "model-00002-of-00002.bs",
202
+ "model.layers.30.mlp.down_proj.weight": "model-00002-of-00002.bs",
203
+ "model.layers.30.mlp.gate_up_proj.weight": "model-00002-of-00002.bs",
204
+ "model.layers.30.post_attention_layernorm.weight": "model-00002-of-00002.bs",
205
+ "model.layers.30.self_attn.o_proj.weight": "model-00002-of-00002.bs",
206
+ "model.layers.30.self_attn.qkv_proj.weight": "model-00002-of-00002.bs",
207
+ "model.layers.31.input_layernorm.weight": "model-00002-of-00002.bs",
208
+ "model.layers.31.mlp.down_proj.weight": "model-00002-of-00002.bs",
209
+ "model.layers.31.mlp.gate_up_proj.weight": "model-00002-of-00002.bs",
210
+ "model.layers.31.post_attention_layernorm.weight": "model-00002-of-00002.bs",
211
+ "model.layers.31.self_attn.o_proj.weight": "model-00002-of-00002.bs",
212
+ "model.layers.31.self_attn.qkv_proj.weight": "model-00002-of-00002.bs",
213
+ "model.norm.weight": "model-00002-of-00002.bs"
214
+ }
215
+ }