Hugging Face's logo

ssdataanalysis
/
AI21-Jamba2-Mini-mlx-8Bit

Text Generation
Transformers
Safetensors
MLX
jamba
conversational
8-bit precision
Model card Files
xet
Community
ssdataanalysis commited on
Commit
c97d35c
·
verified ·
1 Parent(s): 86fd802

Upload config.json with huggingface_hub

Browse files
Files changed (1) hide show
  1. config.json +179 -0
config.json ADDED
@@ -0,0 +1,179 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "JambaForCausalLM"
4
+ ],
5
+ "attention_dropout": 0.0,
6
+ "attn_layer_offset": 4,
7
+ "attn_layer_period": 8,
8
+ "bos_token_id": 1,
9
+ "dtype": "bfloat16",
10
+ "eos_token_id": 519,
11
+ "expert_layer_offset": 1,
12
+ "expert_layer_period": 2,
13
+ "hidden_act": "silu",
14
+ "hidden_size": 4096,
15
+ "initializer_range": 0.02,
16
+ "intermediate_size": 14336,
17
+ "mamba_conv_bias": true,
18
+ "mamba_d_conv": 4,
19
+ "mamba_d_state": 16,
20
+ "mamba_dt_rank": 256,
21
+ "mamba_expand": 2,
22
+ "mamba_proj_bias": false,
23
+ "max_position_embeddings": 262144,
24
+ "model_type": "jamba",
25
+ "num_attention_heads": 32,
26
+ "num_experts": 16,
27
+ "num_experts_per_tok": 2,
28
+ "num_hidden_layers": 32,
29
+ "num_key_value_heads": 8,
30
+ "num_logits_to_keep": 1,
31
+ "output_router_logits": false,
32
+ "pad_token_id": 0,
33
+ "quantization": {
34
+ "group_size": 64,
35
+ "bits": 8,
36
+ "mode": "affine",
37
+ "model.layers.1.feed_forward.router": {
38
+ "group_size": 64,
39
+ "bits": 8
40
+ },
41
+ "model.layers.3.feed_forward.router": {
42
+ "group_size": 64,
43
+ "bits": 8
44
+ },
45
+ "model.layers.5.feed_forward.router": {
46
+ "group_size": 64,
47
+ "bits": 8
48
+ },
49
+ "model.layers.7.feed_forward.router": {
50
+ "group_size": 64,
51
+ "bits": 8
52
+ },
53
+ "model.layers.9.feed_forward.router": {
54
+ "group_size": 64,
55
+ "bits": 8
56
+ },
57
+ "model.layers.11.feed_forward.router": {
58
+ "group_size": 64,
59
+ "bits": 8
60
+ },
61
+ "model.layers.13.feed_forward.router": {
62
+ "group_size": 64,
63
+ "bits": 8
64
+ },
65
+ "model.layers.15.feed_forward.router": {
66
+ "group_size": 64,
67
+ "bits": 8
68
+ },
69
+ "model.layers.17.feed_forward.router": {
70
+ "group_size": 64,
71
+ "bits": 8
72
+ },
73
+ "model.layers.19.feed_forward.router": {
74
+ "group_size": 64,
75
+ "bits": 8
76
+ },
77
+ "model.layers.21.feed_forward.router": {
78
+ "group_size": 64,
79
+ "bits": 8
80
+ },
81
+ "model.layers.23.feed_forward.router": {
82
+ "group_size": 64,
83
+ "bits": 8
84
+ },
85
+ "model.layers.25.feed_forward.router": {
86
+ "group_size": 64,
87
+ "bits": 8
88
+ },
89
+ "model.layers.27.feed_forward.router": {
90
+ "group_size": 64,
91
+ "bits": 8
92
+ },
93
+ "model.layers.29.feed_forward.router": {
94
+ "group_size": 64,
95
+ "bits": 8
96
+ },
97
+ "model.layers.31.feed_forward.router": {
98
+ "group_size": 64,
99
+ "bits": 8
100
+ }
101
+ },
102
+ "quantization_config": {
103
+ "group_size": 64,
104
+ "bits": 8,
105
+ "mode": "affine",
106
+ "model.layers.1.feed_forward.router": {
107
+ "group_size": 64,
108
+ "bits": 8
109
+ },
110
+ "model.layers.3.feed_forward.router": {
111
+ "group_size": 64,
112
+ "bits": 8
113
+ },
114
+ "model.layers.5.feed_forward.router": {
115
+ "group_size": 64,
116
+ "bits": 8
117
+ },
118
+ "model.layers.7.feed_forward.router": {
119
+ "group_size": 64,
120
+ "bits": 8
121
+ },
122
+ "model.layers.9.feed_forward.router": {
123
+ "group_size": 64,
124
+ "bits": 8
125
+ },
126
+ "model.layers.11.feed_forward.router": {
127
+ "group_size": 64,
128
+ "bits": 8
129
+ },
130
+ "model.layers.13.feed_forward.router": {
131
+ "group_size": 64,
132
+ "bits": 8
133
+ },
134
+ "model.layers.15.feed_forward.router": {
135
+ "group_size": 64,
136
+ "bits": 8
137
+ },
138
+ "model.layers.17.feed_forward.router": {
139
+ "group_size": 64,
140
+ "bits": 8
141
+ },
142
+ "model.layers.19.feed_forward.router": {
143
+ "group_size": 64,
144
+ "bits": 8
145
+ },
146
+ "model.layers.21.feed_forward.router": {
147
+ "group_size": 64,
148
+ "bits": 8
149
+ },
150
+ "model.layers.23.feed_forward.router": {
151
+ "group_size": 64,
152
+ "bits": 8
153
+ },
154
+ "model.layers.25.feed_forward.router": {
155
+ "group_size": 64,
156
+ "bits": 8
157
+ },
158
+ "model.layers.27.feed_forward.router": {
159
+ "group_size": 64,
160
+ "bits": 8
161
+ },
162
+ "model.layers.29.feed_forward.router": {
163
+ "group_size": 64,
164
+ "bits": 8
165
+ },
166
+ "model.layers.31.feed_forward.router": {
167
+ "group_size": 64,
168
+ "bits": 8
169
+ }
170
+ },
171
+ "rms_norm_eps": 1e-06,
172
+ "router_aux_loss_coef": 0.001,
173
+ "sliding_window": null,
174
+ "tie_word_embeddings": false,
175
+ "transformers_version": "4.56.1",
176
+ "use_cache": true,
177
+ "use_mamba_kernels": true,
178
+ "vocab_size": 65536
179
+ }