Hugging Face's logo

hermitdave
/
Agnes-3.0-Flash-MLX-6bit-MTP

Text Generation
MLX
Safetensors
English
Chinese
qwen3_5
agnes
6bit
hybrid-attention
gated-delta-net
mtp
conversational
6-bit
Model card Files
xet
 Community
hermitdave commited on
Commit
6cffdf8
·
verified ·
1 Parent(s): 9d0cdcf

Upload folder using huggingface_hub

Browse files
Files changed (3) hide show
  1. config.json +168 -137
  2. model-00005-of-00005.safetensors +2 -2
  3. model.safetensors.index.json +0 -0
config.json CHANGED
@@ -1,142 +1,173 @@
1
  {
2
- "architectures": [
3
- "Qwen3_5ForConditionalGeneration"
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
4
  ],
5
- "eos_token_id": [
6
- 248046,
7
- 248044
8
- ],
9
- "language_model_only": false,
10
- "model_type": "qwen3_5",
11
- "quantization": {
12
- "group_size": 64,
13
- "bits": 6,
14
- "mode": "affine"
15
- },
16
- "quantization_config": {
17
- "group_size": 64,
18
- "bits": 6,
19
- "mode": "affine"
20
- },
21
- "text_config": {
22
- "model_type": "qwen3_5_text",
23
- "hidden_size": 5120,
24
- "num_hidden_layers": 72,
25
- "intermediate_size": 19456,
26
- "vocab_size": 248320,
27
- "layer_types": [
28
- "linear_attention",
29
- "linear_attention",
30
- "linear_attention",
31
- "full_attention",
32
- "linear_attention",
33
- "linear_attention",
34
- "linear_attention",
35
- "full_attention",
36
- "linear_attention",
37
- "linear_attention",
38
- "linear_attention",
39
- "full_attention",
40
- "linear_attention",
41
- "linear_attention",
42
- "linear_attention",
43
- "full_attention",
44
- "linear_attention",
45
- "linear_attention",
46
- "linear_attention",
47
- "full_attention",
48
- "linear_attention",
49
- "linear_attention",
50
- "linear_attention",
51
- "full_attention",
52
- "linear_attention",
53
- "linear_attention",
54
- "linear_attention",
55
- "full_attention",
56
- "linear_attention",
57
- "linear_attention",
58
- "linear_attention",
59
- "full_attention",
60
- "linear_attention",
61
- "linear_attention",
62
- "linear_attention",
63
- "full_attention",
64
- "linear_attention",
65
- "linear_attention",
66
- "linear_attention",
67
- "full_attention",
68
- "linear_attention",
69
- "linear_attention",
70
- "linear_attention",
71
- "full_attention",
72
- "linear_attention",
73
- "linear_attention",
74
- "linear_attention",
75
- "full_attention",
76
- "linear_attention",
77
- "linear_attention",
78
- "linear_attention",
79
- "full_attention",
80
- "linear_attention",
81
- "linear_attention",
82
- "linear_attention",
83
- "full_attention",
84
- "linear_attention",
85
- "linear_attention",
86
- "linear_attention",
87
- "full_attention",
88
- "linear_attention",
89
- "linear_attention",
90
- "linear_attention",
91
- "full_attention",
92
- "linear_attention",
93
- "linear_attention",
94
- "linear_attention",
95
- "full_attention",
96
- "linear_attention",
97
- "linear_attention",
98
- "linear_attention",
99
- "full_attention"
100
- ],
101
- "num_attention_heads": 24,
102
- "num_key_value_heads": 4,
103
- "head_dim": 256,
104
- "attention_bias": false,
105
- "attention_dropout": 0.0,
106
- "attn_output_gate": true,
107
- "partial_rotary_factor": 0.25,
108
- "linear_num_key_heads": 16,
109
- "linear_num_value_heads": 48,
110
- "linear_key_head_dim": 128,
111
- "linear_value_head_dim": 128,
112
- "linear_conv_kernel_dim": 4,
113
- "output_gate_type": "swish",
114
- "mamba_ssm_dtype": "float32",
115
- "hidden_act": "silu",
116
- "rms_norm_eps": 1e-06,
117
- "initializer_range": 0.02,
118
- "rope_parameters": {
119
- "mrope_interleaved": true,
120
- "mrope_section": [
121
- 11,
122
- 11,
123
- 10
124
- ],
125
- "partial_rotary_factor": 0.25,
126
- "rope_theta": 10000000,
127
- "type": "default"
128
- },
129
- "max_position_embeddings": 262144,
130
- "mtp_num_hidden_layers": 0,
131
- "mtp_use_dedicated_embeddings": false,
132
- "bos_token_id": 248044,
133
- "eos_token_id": 248044,
134
- "pad_token_id": null,
135
- "tie_word_embeddings": false,
136
- "use_cache": true,
137
- "dtype": "bfloat16",
138
- "full_attention_interval": 4
139
  },
 
 
 
 
 
 
140
  "tie_word_embeddings": false,
141
- "transformers_version": "5.8.0"
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
142
  }
 
1
  {
2
+ "architectures": [
3
+ "Qwen3_5ForConditionalGeneration"
4
+ ],
5
+ "eos_token_id": [
6
+ 248046,
7
+ 248044
8
+ ],
9
+ "language_model_only": false,
10
+ "model_type": "qwen3_5",
11
+ "quantization": {
12
+ "group_size": 64,
13
+ "bits": 6,
14
+ "mode": "affine"
15
+ },
16
+ "quantization_config": {
17
+ "group_size": 64,
18
+ "bits": 6,
19
+ "mode": "affine"
20
+ },
21
+ "text_config": {
22
+ "model_type": "qwen3_5_text",
23
+ "hidden_size": 5120,
24
+ "num_hidden_layers": 72,
25
+ "intermediate_size": 19456,
26
+ "vocab_size": 248320,
27
+ "layer_types": [
28
+ "linear_attention",
29
+ "linear_attention",
30
+ "linear_attention",
31
+ "full_attention",
32
+ "linear_attention",
33
+ "linear_attention",
34
+ "linear_attention",
35
+ "full_attention",
36
+ "linear_attention",
37
+ "linear_attention",
38
+ "linear_attention",
39
+ "full_attention",
40
+ "linear_attention",
41
+ "linear_attention",
42
+ "linear_attention",
43
+ "full_attention",
44
+ "linear_attention",
45
+ "linear_attention",
46
+ "linear_attention",
47
+ "full_attention",
48
+ "linear_attention",
49
+ "linear_attention",
50
+ "linear_attention",
51
+ "full_attention",
52
+ "linear_attention",
53
+ "linear_attention",
54
+ "linear_attention",
55
+ "full_attention",
56
+ "linear_attention",
57
+ "linear_attention",
58
+ "linear_attention",
59
+ "full_attention",
60
+ "linear_attention",
61
+ "linear_attention",
62
+ "linear_attention",
63
+ "full_attention",
64
+ "linear_attention",
65
+ "linear_attention",
66
+ "linear_attention",
67
+ "full_attention",
68
+ "linear_attention",
69
+ "linear_attention",
70
+ "linear_attention",
71
+ "full_attention",
72
+ "linear_attention",
73
+ "linear_attention",
74
+ "linear_attention",
75
+ "full_attention",
76
+ "linear_attention",
77
+ "linear_attention",
78
+ "linear_attention",
79
+ "full_attention",
80
+ "linear_attention",
81
+ "linear_attention",
82
+ "linear_attention",
83
+ "full_attention",
84
+ "linear_attention",
85
+ "linear_attention",
86
+ "linear_attention",
87
+ "full_attention",
88
+ "linear_attention",
89
+ "linear_attention",
90
+ "linear_attention",
91
+ "full_attention",
92
+ "linear_attention",
93
+ "linear_attention",
94
+ "linear_attention",
95
+ "full_attention",
96
+ "linear_attention",
97
+ "linear_attention",
98
+ "linear_attention",
99
+ "full_attention"
100
  ],
101
+ "num_attention_heads": 24,
102
+ "num_key_value_heads": 4,
103
+ "head_dim": 256,
104
+ "attention_bias": false,
105
+ "attention_dropout": 0.0,
106
+ "attn_output_gate": true,
107
+ "partial_rotary_factor": 0.25,
108
+ "linear_num_key_heads": 16,
109
+ "linear_num_value_heads": 48,
110
+ "linear_key_head_dim": 128,
111
+ "linear_value_head_dim": 128,
112
+ "linear_conv_kernel_dim": 4,
113
+ "output_gate_type": "swish",
114
+ "mamba_ssm_dtype": "float32",
115
+ "hidden_act": "silu",
116
+ "rms_norm_eps": 1e-06,
117
+ "initializer_range": 0.02,
118
+ "rope_parameters": {
119
+ "mrope_interleaved": true,
120
+ "mrope_section": [
121
+ 11,
122
+ 11,
123
+ 10
124
+ ],
125
+ "partial_rotary_factor": 0.25,
126
+ "rope_theta": 10000000,
127
+ "type": "default"
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
128
  },
129
+ "max_position_embeddings": 262144,
130
+ "mtp_num_hidden_layers": 1,
131
+ "mtp_use_dedicated_embeddings": false,
132
+ "bos_token_id": 248044,
133
+ "eos_token_id": 248044,
134
+ "pad_token_id": null,
135
  "tie_word_embeddings": false,
136
+ "use_cache": true,
137
+ "dtype": "bfloat16",
138
+ "full_attention_interval": 4
139
+ },
140
+ "tie_word_embeddings": false,
141
+ "transformers_version": "5.8.0",
142
+ "vision_config": {
143
+ "deepstack_visual_indexes": [],
144
+ "depth": 27,
145
+ "hidden_act": "gelu_pytorch_tanh",
146
+ "hidden_size": 1152,
147
+ "in_channels": 3,
148
+ "initializer_range": 0.02,
149
+ "intermediate_size": 4304,
150
+ "model_type": "qwen3_5",
151
+ "num_heads": 16,
152
+ "num_position_embeddings": 2304,
153
+ "out_hidden_size": 5120,
154
+ "patch_size": 16,
155
+ "spatial_merge_size": 2,
156
+ "temporal_patch_size": 2
157
+ },
158
+ "image_token_id": 248056,
159
+ "video_token_id": 248057,
160
+ "vision_start_token_id": 248053,
161
+ "vision_end_token_id": 248054,
162
+ "mtp_num_hidden_layers": 1,
163
+ "mtp": {
164
+ "hidden_size": 5120,
165
+ "intermediate_size": 17408,
166
+ "num_hidden_layers": 0,
167
+ "num_attention_heads": 24,
168
+ "num_key_value_heads": 4,
169
+ "head_dim": 256,
170
+ "mtp_num_hidden_layers": 1,
171
+ "mtp_use_dedicated_embeddings": false
172
+ }
173
  }
model-00005-of-00005.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:eafa69833220d57daadc76620f331dc11b2bf285dc1b4b5ce820b543646b3f63
3
- size 4794324583
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dd093581e9ce43aad0553ac0ec843226c746b9ae14c8f10ceef9b959a823da85
3
+ size 1253492949
model.safetensors.index.json CHANGED
The diff for this file is too large to render. See raw diff