mihainadas commited on
Commit
0bb6b79
·
verified ·
1 Parent(s): b5702a6

Upload folder using huggingface_hub

Browse files
config.json ADDED
@@ -0,0 +1,33 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "MambaForCausalLM"
4
+ ],
5
+ "bos_token_id": 2,
6
+ "conv_kernel": 4,
7
+ "dtype": "float32",
8
+ "eos_token_id": 3,
9
+ "expand": 2,
10
+ "hidden_act": "silu",
11
+ "hidden_size": 512,
12
+ "initializer_range": 0.1,
13
+ "intermediate_size": 1024,
14
+ "layer_norm_epsilon": 1e-05,
15
+ "model_type": "mamba",
16
+ "num_hidden_layers": 18,
17
+ "pad_token_id": 0,
18
+ "rescale_prenorm_residual": false,
19
+ "residual_in_fp32": true,
20
+ "state_size": 64,
21
+ "time_step_floor": 0.0001,
22
+ "time_step_init_scheme": "random",
23
+ "time_step_max": 0.1,
24
+ "time_step_min": 0.001,
25
+ "time_step_rank": 32,
26
+ "time_step_scale": 1.0,
27
+ "transformers_version": "4.56.1",
28
+ "use_bias": false,
29
+ "use_cache": true,
30
+ "use_conv_bias": true,
31
+ "use_mambapy": false,
32
+ "vocab_size": 32000
33
+ }
generation_config.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 2,
4
+ "eos_token_id": [
5
+ 3
6
+ ],
7
+ "pad_token_id": 0,
8
+ "transformers_version": "4.56.1"
9
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6474c768ebe6fc5e595a71f3ff0f9a2a264b7633423bca7ea09a2a79869e104a
3
+ size 198230947
model.safetensors.index.json ADDED
@@ -0,0 +1,190 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "metadata": {
3
+ "total_size": 198211584,
4
+ "total_parameters": 48354816
5
+ },
6
+ "weight_map": {
7
+ "backbone.embeddings.weight": "model.safetensors",
8
+ "backbone.layers.0.mixer.A_log": "model.safetensors",
9
+ "backbone.layers.0.mixer.D": "model.safetensors",
10
+ "backbone.layers.0.mixer.conv1d.bias": "model.safetensors",
11
+ "backbone.layers.0.mixer.conv1d.weight": "model.safetensors",
12
+ "backbone.layers.0.mixer.dt_proj.bias": "model.safetensors",
13
+ "backbone.layers.0.mixer.dt_proj.weight": "model.safetensors",
14
+ "backbone.layers.0.mixer.in_proj.weight": "model.safetensors",
15
+ "backbone.layers.0.mixer.out_proj.weight": "model.safetensors",
16
+ "backbone.layers.0.mixer.x_proj.weight": "model.safetensors",
17
+ "backbone.layers.0.norm.weight": "model.safetensors",
18
+ "backbone.layers.1.mixer.A_log": "model.safetensors",
19
+ "backbone.layers.1.mixer.D": "model.safetensors",
20
+ "backbone.layers.1.mixer.conv1d.bias": "model.safetensors",
21
+ "backbone.layers.1.mixer.conv1d.weight": "model.safetensors",
22
+ "backbone.layers.1.mixer.dt_proj.bias": "model.safetensors",
23
+ "backbone.layers.1.mixer.dt_proj.weight": "model.safetensors",
24
+ "backbone.layers.1.mixer.in_proj.weight": "model.safetensors",
25
+ "backbone.layers.1.mixer.out_proj.weight": "model.safetensors",
26
+ "backbone.layers.1.mixer.x_proj.weight": "model.safetensors",
27
+ "backbone.layers.1.norm.weight": "model.safetensors",
28
+ "backbone.layers.10.mixer.A_log": "model.safetensors",
29
+ "backbone.layers.10.mixer.D": "model.safetensors",
30
+ "backbone.layers.10.mixer.conv1d.bias": "model.safetensors",
31
+ "backbone.layers.10.mixer.conv1d.weight": "model.safetensors",
32
+ "backbone.layers.10.mixer.dt_proj.bias": "model.safetensors",
33
+ "backbone.layers.10.mixer.dt_proj.weight": "model.safetensors",
34
+ "backbone.layers.10.mixer.in_proj.weight": "model.safetensors",
35
+ "backbone.layers.10.mixer.out_proj.weight": "model.safetensors",
36
+ "backbone.layers.10.mixer.x_proj.weight": "model.safetensors",
37
+ "backbone.layers.10.norm.weight": "model.safetensors",
38
+ "backbone.layers.11.mixer.A_log": "model.safetensors",
39
+ "backbone.layers.11.mixer.D": "model.safetensors",
40
+ "backbone.layers.11.mixer.conv1d.bias": "model.safetensors",
41
+ "backbone.layers.11.mixer.conv1d.weight": "model.safetensors",
42
+ "backbone.layers.11.mixer.dt_proj.bias": "model.safetensors",
43
+ "backbone.layers.11.mixer.dt_proj.weight": "model.safetensors",
44
+ "backbone.layers.11.mixer.in_proj.weight": "model.safetensors",
45
+ "backbone.layers.11.mixer.out_proj.weight": "model.safetensors",
46
+ "backbone.layers.11.mixer.x_proj.weight": "model.safetensors",
47
+ "backbone.layers.11.norm.weight": "model.safetensors",
48
+ "backbone.layers.12.mixer.A_log": "model.safetensors",
49
+ "backbone.layers.12.mixer.D": "model.safetensors",
50
+ "backbone.layers.12.mixer.conv1d.bias": "model.safetensors",
51
+ "backbone.layers.12.mixer.conv1d.weight": "model.safetensors",
52
+ "backbone.layers.12.mixer.dt_proj.bias": "model.safetensors",
53
+ "backbone.layers.12.mixer.dt_proj.weight": "model.safetensors",
54
+ "backbone.layers.12.mixer.in_proj.weight": "model.safetensors",
55
+ "backbone.layers.12.mixer.out_proj.weight": "model.safetensors",
56
+ "backbone.layers.12.mixer.x_proj.weight": "model.safetensors",
57
+ "backbone.layers.12.norm.weight": "model.safetensors",
58
+ "backbone.layers.13.mixer.A_log": "model.safetensors",
59
+ "backbone.layers.13.mixer.D": "model.safetensors",
60
+ "backbone.layers.13.mixer.conv1d.bias": "model.safetensors",
61
+ "backbone.layers.13.mixer.conv1d.weight": "model.safetensors",
62
+ "backbone.layers.13.mixer.dt_proj.bias": "model.safetensors",
63
+ "backbone.layers.13.mixer.dt_proj.weight": "model.safetensors",
64
+ "backbone.layers.13.mixer.in_proj.weight": "model.safetensors",
65
+ "backbone.layers.13.mixer.out_proj.weight": "model.safetensors",
66
+ "backbone.layers.13.mixer.x_proj.weight": "model.safetensors",
67
+ "backbone.layers.13.norm.weight": "model.safetensors",
68
+ "backbone.layers.14.mixer.A_log": "model.safetensors",
69
+ "backbone.layers.14.mixer.D": "model.safetensors",
70
+ "backbone.layers.14.mixer.conv1d.bias": "model.safetensors",
71
+ "backbone.layers.14.mixer.conv1d.weight": "model.safetensors",
72
+ "backbone.layers.14.mixer.dt_proj.bias": "model.safetensors",
73
+ "backbone.layers.14.mixer.dt_proj.weight": "model.safetensors",
74
+ "backbone.layers.14.mixer.in_proj.weight": "model.safetensors",
75
+ "backbone.layers.14.mixer.out_proj.weight": "model.safetensors",
76
+ "backbone.layers.14.mixer.x_proj.weight": "model.safetensors",
77
+ "backbone.layers.14.norm.weight": "model.safetensors",
78
+ "backbone.layers.15.mixer.A_log": "model.safetensors",
79
+ "backbone.layers.15.mixer.D": "model.safetensors",
80
+ "backbone.layers.15.mixer.conv1d.bias": "model.safetensors",
81
+ "backbone.layers.15.mixer.conv1d.weight": "model.safetensors",
82
+ "backbone.layers.15.mixer.dt_proj.bias": "model.safetensors",
83
+ "backbone.layers.15.mixer.dt_proj.weight": "model.safetensors",
84
+ "backbone.layers.15.mixer.in_proj.weight": "model.safetensors",
85
+ "backbone.layers.15.mixer.out_proj.weight": "model.safetensors",
86
+ "backbone.layers.15.mixer.x_proj.weight": "model.safetensors",
87
+ "backbone.layers.15.norm.weight": "model.safetensors",
88
+ "backbone.layers.16.mixer.A_log": "model.safetensors",
89
+ "backbone.layers.16.mixer.D": "model.safetensors",
90
+ "backbone.layers.16.mixer.conv1d.bias": "model.safetensors",
91
+ "backbone.layers.16.mixer.conv1d.weight": "model.safetensors",
92
+ "backbone.layers.16.mixer.dt_proj.bias": "model.safetensors",
93
+ "backbone.layers.16.mixer.dt_proj.weight": "model.safetensors",
94
+ "backbone.layers.16.mixer.in_proj.weight": "model.safetensors",
95
+ "backbone.layers.16.mixer.out_proj.weight": "model.safetensors",
96
+ "backbone.layers.16.mixer.x_proj.weight": "model.safetensors",
97
+ "backbone.layers.16.norm.weight": "model.safetensors",
98
+ "backbone.layers.17.mixer.A_log": "model.safetensors",
99
+ "backbone.layers.17.mixer.D": "model.safetensors",
100
+ "backbone.layers.17.mixer.conv1d.bias": "model.safetensors",
101
+ "backbone.layers.17.mixer.conv1d.weight": "model.safetensors",
102
+ "backbone.layers.17.mixer.dt_proj.bias": "model.safetensors",
103
+ "backbone.layers.17.mixer.dt_proj.weight": "model.safetensors",
104
+ "backbone.layers.17.mixer.in_proj.weight": "model.safetensors",
105
+ "backbone.layers.17.mixer.out_proj.weight": "model.safetensors",
106
+ "backbone.layers.17.mixer.x_proj.weight": "model.safetensors",
107
+ "backbone.layers.17.norm.weight": "model.safetensors",
108
+ "backbone.layers.2.mixer.A_log": "model.safetensors",
109
+ "backbone.layers.2.mixer.D": "model.safetensors",
110
+ "backbone.layers.2.mixer.conv1d.bias": "model.safetensors",
111
+ "backbone.layers.2.mixer.conv1d.weight": "model.safetensors",
112
+ "backbone.layers.2.mixer.dt_proj.bias": "model.safetensors",
113
+ "backbone.layers.2.mixer.dt_proj.weight": "model.safetensors",
114
+ "backbone.layers.2.mixer.in_proj.weight": "model.safetensors",
115
+ "backbone.layers.2.mixer.out_proj.weight": "model.safetensors",
116
+ "backbone.layers.2.mixer.x_proj.weight": "model.safetensors",
117
+ "backbone.layers.2.norm.weight": "model.safetensors",
118
+ "backbone.layers.3.mixer.A_log": "model.safetensors",
119
+ "backbone.layers.3.mixer.D": "model.safetensors",
120
+ "backbone.layers.3.mixer.conv1d.bias": "model.safetensors",
121
+ "backbone.layers.3.mixer.conv1d.weight": "model.safetensors",
122
+ "backbone.layers.3.mixer.dt_proj.bias": "model.safetensors",
123
+ "backbone.layers.3.mixer.dt_proj.weight": "model.safetensors",
124
+ "backbone.layers.3.mixer.in_proj.weight": "model.safetensors",
125
+ "backbone.layers.3.mixer.out_proj.weight": "model.safetensors",
126
+ "backbone.layers.3.mixer.x_proj.weight": "model.safetensors",
127
+ "backbone.layers.3.norm.weight": "model.safetensors",
128
+ "backbone.layers.4.mixer.A_log": "model.safetensors",
129
+ "backbone.layers.4.mixer.D": "model.safetensors",
130
+ "backbone.layers.4.mixer.conv1d.bias": "model.safetensors",
131
+ "backbone.layers.4.mixer.conv1d.weight": "model.safetensors",
132
+ "backbone.layers.4.mixer.dt_proj.bias": "model.safetensors",
133
+ "backbone.layers.4.mixer.dt_proj.weight": "model.safetensors",
134
+ "backbone.layers.4.mixer.in_proj.weight": "model.safetensors",
135
+ "backbone.layers.4.mixer.out_proj.weight": "model.safetensors",
136
+ "backbone.layers.4.mixer.x_proj.weight": "model.safetensors",
137
+ "backbone.layers.4.norm.weight": "model.safetensors",
138
+ "backbone.layers.5.mixer.A_log": "model.safetensors",
139
+ "backbone.layers.5.mixer.D": "model.safetensors",
140
+ "backbone.layers.5.mixer.conv1d.bias": "model.safetensors",
141
+ "backbone.layers.5.mixer.conv1d.weight": "model.safetensors",
142
+ "backbone.layers.5.mixer.dt_proj.bias": "model.safetensors",
143
+ "backbone.layers.5.mixer.dt_proj.weight": "model.safetensors",
144
+ "backbone.layers.5.mixer.in_proj.weight": "model.safetensors",
145
+ "backbone.layers.5.mixer.out_proj.weight": "model.safetensors",
146
+ "backbone.layers.5.mixer.x_proj.weight": "model.safetensors",
147
+ "backbone.layers.5.norm.weight": "model.safetensors",
148
+ "backbone.layers.6.mixer.A_log": "model.safetensors",
149
+ "backbone.layers.6.mixer.D": "model.safetensors",
150
+ "backbone.layers.6.mixer.conv1d.bias": "model.safetensors",
151
+ "backbone.layers.6.mixer.conv1d.weight": "model.safetensors",
152
+ "backbone.layers.6.mixer.dt_proj.bias": "model.safetensors",
153
+ "backbone.layers.6.mixer.dt_proj.weight": "model.safetensors",
154
+ "backbone.layers.6.mixer.in_proj.weight": "model.safetensors",
155
+ "backbone.layers.6.mixer.out_proj.weight": "model.safetensors",
156
+ "backbone.layers.6.mixer.x_proj.weight": "model.safetensors",
157
+ "backbone.layers.6.norm.weight": "model.safetensors",
158
+ "backbone.layers.7.mixer.A_log": "model.safetensors",
159
+ "backbone.layers.7.mixer.D": "model.safetensors",
160
+ "backbone.layers.7.mixer.conv1d.bias": "model.safetensors",
161
+ "backbone.layers.7.mixer.conv1d.weight": "model.safetensors",
162
+ "backbone.layers.7.mixer.dt_proj.bias": "model.safetensors",
163
+ "backbone.layers.7.mixer.dt_proj.weight": "model.safetensors",
164
+ "backbone.layers.7.mixer.in_proj.weight": "model.safetensors",
165
+ "backbone.layers.7.mixer.out_proj.weight": "model.safetensors",
166
+ "backbone.layers.7.mixer.x_proj.weight": "model.safetensors",
167
+ "backbone.layers.7.norm.weight": "model.safetensors",
168
+ "backbone.layers.8.mixer.A_log": "model.safetensors",
169
+ "backbone.layers.8.mixer.D": "model.safetensors",
170
+ "backbone.layers.8.mixer.conv1d.bias": "model.safetensors",
171
+ "backbone.layers.8.mixer.conv1d.weight": "model.safetensors",
172
+ "backbone.layers.8.mixer.dt_proj.bias": "model.safetensors",
173
+ "backbone.layers.8.mixer.dt_proj.weight": "model.safetensors",
174
+ "backbone.layers.8.mixer.in_proj.weight": "model.safetensors",
175
+ "backbone.layers.8.mixer.out_proj.weight": "model.safetensors",
176
+ "backbone.layers.8.mixer.x_proj.weight": "model.safetensors",
177
+ "backbone.layers.8.norm.weight": "model.safetensors",
178
+ "backbone.layers.9.mixer.A_log": "model.safetensors",
179
+ "backbone.layers.9.mixer.D": "model.safetensors",
180
+ "backbone.layers.9.mixer.conv1d.bias": "model.safetensors",
181
+ "backbone.layers.9.mixer.conv1d.weight": "model.safetensors",
182
+ "backbone.layers.9.mixer.dt_proj.bias": "model.safetensors",
183
+ "backbone.layers.9.mixer.dt_proj.weight": "model.safetensors",
184
+ "backbone.layers.9.mixer.in_proj.weight": "model.safetensors",
185
+ "backbone.layers.9.mixer.out_proj.weight": "model.safetensors",
186
+ "backbone.layers.9.mixer.x_proj.weight": "model.safetensors",
187
+ "backbone.layers.9.norm.weight": "model.safetensors",
188
+ "backbone.norm_f.weight": "model.safetensors"
189
+ }
190
+ }
special_tokens_map.json ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<bos>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "<eos>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "<pad>",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "unk_token": {
24
+ "content": "<unk>",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ }
30
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,44 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "<pad>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "<unk>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "<bos>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "<eos>",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ }
35
+ },
36
+ "bos_token": "<bos>",
37
+ "clean_up_tokenization_spaces": false,
38
+ "eos_token": "<eos>",
39
+ "extra_special_tokens": {},
40
+ "model_max_length": 1000000000000000019884624838656,
41
+ "pad_token": "<pad>",
42
+ "tokenizer_class": "PreTrainedTokenizerFast",
43
+ "unk_token": "<unk>"
44
+ }