Hugging Face's logo

cyankiwi
/
Devstral-Small-2-24B-Instruct-2512-AWQ-4bit

Safetensors
vllm
mistral3
mistral-common
compressed-tensors
Model card Files
xet
Community
3
cpatonn commited on
Commit
da6366e
·
verified ·
1 Parent(s): f20dde7

Delete files *ipynb* with huggingface_hub

Browse files
Files changed (2) hide show
  1. .ipynb_checkpoints/config-checkpoint.json +0 -267
  2. .ipynb_checkpoints/model.safetensors.index-checkpoint.json +0 -0
.ipynb_checkpoints/config-checkpoint.json DELETED
@@ -1,267 +0,0 @@
1
- {
2
- "architectures": [
3
- "Mistral3ForConditionalGeneration"
4
- ],
5
- "dtype": "bfloat16",
6
- "image_token_index": 10,
7
- "model_type": "mistral3",
8
- "multimodal_projector_bias": false,
9
- "projector_hidden_act": "gelu",
10
- "tie_word_embeddings": false,
11
- "quantization_config": {
12
- "config_groups": {
13
- "group_0": {
14
- "format": "pack-quantized",
15
- "input_activations": null,
16
- "output_activations": null,
17
- "targets": [
18
- "Linear"
19
- ],
20
- "weights": {
21
- "actorder": null,
22
- "block_structure": null,
23
- "dynamic": false,
24
- "group_size": 32,
25
- "num_bits": 4,
26
- "observer": "mse",
27
- "observer_kwargs": {},
28
- "strategy": "group",
29
- "symmetric": true,
30
- "type": "int"
31
- }
32
- }
33
- },
34
- "format": "pack-quantized",
35
- "global_compression_ratio": null,
36
- "ignore": [
37
- "model.vision_tower.transformer.layers.0.feed_forward.gate_proj",
38
- "model.vision_tower.transformer.layers.0.feed_forward.up_proj",
39
- "model.vision_tower.transformer.layers.0.feed_forward.down_proj",
40
- "model.vision_tower.transformer.layers.0.attention.k_proj",
41
- "model.vision_tower.transformer.layers.0.attention.v_proj",
42
- "model.vision_tower.transformer.layers.0.attention.q_proj",
43
- "model.vision_tower.transformer.layers.0.attention.o_proj",
44
- "model.vision_tower.transformer.layers.1.feed_forward.gate_proj",
45
- "model.vision_tower.transformer.layers.1.feed_forward.up_proj",
46
- "model.vision_tower.transformer.layers.1.feed_forward.down_proj",
47
- "model.vision_tower.transformer.layers.1.attention.k_proj",
48
- "model.vision_tower.transformer.layers.1.attention.v_proj",
49
- "model.vision_tower.transformer.layers.1.attention.q_proj",
50
- "model.vision_tower.transformer.layers.1.attention.o_proj",
51
- "model.vision_tower.transformer.layers.2.feed_forward.gate_proj",
52
- "model.vision_tower.transformer.layers.2.feed_forward.up_proj",
53
- "model.vision_tower.transformer.layers.2.feed_forward.down_proj",
54
- "model.vision_tower.transformer.layers.2.attention.k_proj",
55
- "model.vision_tower.transformer.layers.2.attention.v_proj",
56
- "model.vision_tower.transformer.layers.2.attention.q_proj",
57
- "model.vision_tower.transformer.layers.2.attention.o_proj",
58
- "model.vision_tower.transformer.layers.3.feed_forward.gate_proj",
59
- "model.vision_tower.transformer.layers.3.feed_forward.up_proj",
60
- "model.vision_tower.transformer.layers.3.feed_forward.down_proj",
61
- "model.vision_tower.transformer.layers.3.attention.k_proj",
62
- "model.vision_tower.transformer.layers.3.attention.v_proj",
63
- "model.vision_tower.transformer.layers.3.attention.q_proj",
64
- "model.vision_tower.transformer.layers.3.attention.o_proj",
65
- "model.vision_tower.transformer.layers.4.feed_forward.gate_proj",
66
- "model.vision_tower.transformer.layers.4.feed_forward.up_proj",
67
- "model.vision_tower.transformer.layers.4.feed_forward.down_proj",
68
- "model.vision_tower.transformer.layers.4.attention.k_proj",
69
- "model.vision_tower.transformer.layers.4.attention.v_proj",
70
- "model.vision_tower.transformer.layers.4.attention.q_proj",
71
- "model.vision_tower.transformer.layers.4.attention.o_proj",
72
- "model.vision_tower.transformer.layers.5.feed_forward.gate_proj",
73
- "model.vision_tower.transformer.layers.5.feed_forward.up_proj",
74
- "model.vision_tower.transformer.layers.5.feed_forward.down_proj",
75
- "model.vision_tower.transformer.layers.5.attention.k_proj",
76
- "model.vision_tower.transformer.layers.5.attention.v_proj",
77
- "model.vision_tower.transformer.layers.5.attention.q_proj",
78
- "model.vision_tower.transformer.layers.5.attention.o_proj",
79
- "model.vision_tower.transformer.layers.6.feed_forward.gate_proj",
80
- "model.vision_tower.transformer.layers.6.feed_forward.up_proj",
81
- "model.vision_tower.transformer.layers.6.feed_forward.down_proj",
82
- "model.vision_tower.transformer.layers.6.attention.k_proj",
83
- "model.vision_tower.transformer.layers.6.attention.v_proj",
84
- "model.vision_tower.transformer.layers.6.attention.q_proj",
85
- "model.vision_tower.transformer.layers.6.attention.o_proj",
86
- "model.vision_tower.transformer.layers.7.feed_forward.gate_proj",
87
- "model.vision_tower.transformer.layers.7.feed_forward.up_proj",
88
- "model.vision_tower.transformer.layers.7.feed_forward.down_proj",
89
- "model.vision_tower.transformer.layers.7.attention.k_proj",
90
- "model.vision_tower.transformer.layers.7.attention.v_proj",
91
- "model.vision_tower.transformer.layers.7.attention.q_proj",
92
- "model.vision_tower.transformer.layers.7.attention.o_proj",
93
- "model.vision_tower.transformer.layers.8.feed_forward.gate_proj",
94
- "model.vision_tower.transformer.layers.8.feed_forward.up_proj",
95
- "model.vision_tower.transformer.layers.8.feed_forward.down_proj",
96
- "model.vision_tower.transformer.layers.8.attention.k_proj",
97
- "model.vision_tower.transformer.layers.8.attention.v_proj",
98
- "model.vision_tower.transformer.layers.8.attention.q_proj",
99
- "model.vision_tower.transformer.layers.8.attention.o_proj",
100
- "model.vision_tower.transformer.layers.9.feed_forward.gate_proj",
101
- "model.vision_tower.transformer.layers.9.feed_forward.up_proj",
102
- "model.vision_tower.transformer.layers.9.feed_forward.down_proj",
103
- "model.vision_tower.transformer.layers.9.attention.k_proj",
104
- "model.vision_tower.transformer.layers.9.attention.v_proj",
105
- "model.vision_tower.transformer.layers.9.attention.q_proj",
106
- "model.vision_tower.transformer.layers.9.attention.o_proj",
107
- "model.vision_tower.transformer.layers.10.feed_forward.gate_proj",
108
- "model.vision_tower.transformer.layers.10.feed_forward.up_proj",
109
- "model.vision_tower.transformer.layers.10.feed_forward.down_proj",
110
- "model.vision_tower.transformer.layers.10.attention.k_proj",
111
- "model.vision_tower.transformer.layers.10.attention.v_proj",
112
- "model.vision_tower.transformer.layers.10.attention.q_proj",
113
- "model.vision_tower.transformer.layers.10.attention.o_proj",
114
- "model.vision_tower.transformer.layers.11.feed_forward.gate_proj",
115
- "model.vision_tower.transformer.layers.11.feed_forward.up_proj",
116
- "model.vision_tower.transformer.layers.11.feed_forward.down_proj",
117
- "model.vision_tower.transformer.layers.11.attention.k_proj",
118
- "model.vision_tower.transformer.layers.11.attention.v_proj",
119
- "model.vision_tower.transformer.layers.11.attention.q_proj",
120
- "model.vision_tower.transformer.layers.11.attention.o_proj",
121
- "model.vision_tower.transformer.layers.12.feed_forward.gate_proj",
122
- "model.vision_tower.transformer.layers.12.feed_forward.up_proj",
123
- "model.vision_tower.transformer.layers.12.feed_forward.down_proj",
124
- "model.vision_tower.transformer.layers.12.attention.k_proj",
125
- "model.vision_tower.transformer.layers.12.attention.v_proj",
126
- "model.vision_tower.transformer.layers.12.attention.q_proj",
127
- "model.vision_tower.transformer.layers.12.attention.o_proj",
128
- "model.vision_tower.transformer.layers.13.feed_forward.gate_proj",
129
- "model.vision_tower.transformer.layers.13.feed_forward.up_proj",
130
- "model.vision_tower.transformer.layers.13.feed_forward.down_proj",
131
- "model.vision_tower.transformer.layers.13.attention.k_proj",
132
- "model.vision_tower.transformer.layers.13.attention.v_proj",
133
- "model.vision_tower.transformer.layers.13.attention.q_proj",
134
- "model.vision_tower.transformer.layers.13.attention.o_proj",
135
- "model.vision_tower.transformer.layers.14.feed_forward.gate_proj",
136
- "model.vision_tower.transformer.layers.14.feed_forward.up_proj",
137
- "model.vision_tower.transformer.layers.14.feed_forward.down_proj",
138
- "model.vision_tower.transformer.layers.14.attention.k_proj",
139
- "model.vision_tower.transformer.layers.14.attention.v_proj",
140
- "model.vision_tower.transformer.layers.14.attention.q_proj",
141
- "model.vision_tower.transformer.layers.14.attention.o_proj",
142
- "model.vision_tower.transformer.layers.15.feed_forward.gate_proj",
143
- "model.vision_tower.transformer.layers.15.feed_forward.up_proj",
144
- "model.vision_tower.transformer.layers.15.feed_forward.down_proj",
145
- "model.vision_tower.transformer.layers.15.attention.k_proj",
146
- "model.vision_tower.transformer.layers.15.attention.v_proj",
147
- "model.vision_tower.transformer.layers.15.attention.q_proj",
148
- "model.vision_tower.transformer.layers.15.attention.o_proj",
149
- "model.vision_tower.transformer.layers.16.feed_forward.gate_proj",
150
- "model.vision_tower.transformer.layers.16.feed_forward.up_proj",
151
- "model.vision_tower.transformer.layers.16.feed_forward.down_proj",
152
- "model.vision_tower.transformer.layers.16.attention.k_proj",
153
- "model.vision_tower.transformer.layers.16.attention.v_proj",
154
- "model.vision_tower.transformer.layers.16.attention.q_proj",
155
- "model.vision_tower.transformer.layers.16.attention.o_proj",
156
- "model.vision_tower.transformer.layers.17.feed_forward.gate_proj",
157
- "model.vision_tower.transformer.layers.17.feed_forward.up_proj",
158
- "model.vision_tower.transformer.layers.17.feed_forward.down_proj",
159
- "model.vision_tower.transformer.layers.17.attention.k_proj",
160
- "model.vision_tower.transformer.layers.17.attention.v_proj",
161
- "model.vision_tower.transformer.layers.17.attention.q_proj",
162
- "model.vision_tower.transformer.layers.17.attention.o_proj",
163
- "model.vision_tower.transformer.layers.18.feed_forward.gate_proj",
164
- "model.vision_tower.transformer.layers.18.feed_forward.up_proj",
165
- "model.vision_tower.transformer.layers.18.feed_forward.down_proj",
166
- "model.vision_tower.transformer.layers.18.attention.k_proj",
167
- "model.vision_tower.transformer.layers.18.attention.v_proj",
168
- "model.vision_tower.transformer.layers.18.attention.q_proj",
169
- "model.vision_tower.transformer.layers.18.attention.o_proj",
170
- "model.vision_tower.transformer.layers.19.feed_forward.gate_proj",
171
- "model.vision_tower.transformer.layers.19.feed_forward.up_proj",
172
- "model.vision_tower.transformer.layers.19.feed_forward.down_proj",
173
- "model.vision_tower.transformer.layers.19.attention.k_proj",
174
- "model.vision_tower.transformer.layers.19.attention.v_proj",
175
- "model.vision_tower.transformer.layers.19.attention.q_proj",
176
- "model.vision_tower.transformer.layers.19.attention.o_proj",
177
- "model.vision_tower.transformer.layers.20.feed_forward.gate_proj",
178
- "model.vision_tower.transformer.layers.20.feed_forward.up_proj",
179
- "model.vision_tower.transformer.layers.20.feed_forward.down_proj",
180
- "model.vision_tower.transformer.layers.20.attention.k_proj",
181
- "model.vision_tower.transformer.layers.20.attention.v_proj",
182
- "model.vision_tower.transformer.layers.20.attention.q_proj",
183
- "model.vision_tower.transformer.layers.20.attention.o_proj",
184
- "model.vision_tower.transformer.layers.21.feed_forward.gate_proj",
185
- "model.vision_tower.transformer.layers.21.feed_forward.up_proj",
186
- "model.vision_tower.transformer.layers.21.feed_forward.down_proj",
187
- "model.vision_tower.transformer.layers.21.attention.k_proj",
188
- "model.vision_tower.transformer.layers.21.attention.v_proj",
189
- "model.vision_tower.transformer.layers.21.attention.q_proj",
190
- "model.vision_tower.transformer.layers.21.attention.o_proj",
191
- "model.vision_tower.transformer.layers.22.feed_forward.gate_proj",
192
- "model.vision_tower.transformer.layers.22.feed_forward.up_proj",
193
- "model.vision_tower.transformer.layers.22.feed_forward.down_proj",
194
- "model.vision_tower.transformer.layers.22.attention.k_proj",
195
- "model.vision_tower.transformer.layers.22.attention.v_proj",
196
- "model.vision_tower.transformer.layers.22.attention.q_proj",
197
- "model.vision_tower.transformer.layers.22.attention.o_proj",
198
- "model.vision_tower.transformer.layers.23.feed_forward.gate_proj",
199
- "model.vision_tower.transformer.layers.23.feed_forward.up_proj",
200
- "model.vision_tower.transformer.layers.23.feed_forward.down_proj",
201
- "model.vision_tower.transformer.layers.23.attention.k_proj",
202
- "model.vision_tower.transformer.layers.23.attention.v_proj",
203
- "model.vision_tower.transformer.layers.23.attention.q_proj",
204
- "model.vision_tower.transformer.layers.23.attention.o_proj",
205
- "model.multi_modal_projector.patch_merger.merging_layer",
206
- "model.multi_modal_projector.linear_1",
207
- "model.multi_modal_projector.linear_2",
208
- "lm_head"
209
- ],
210
- "kv_cache_scheme": null,
211
- "quant_method": "compressed-tensors",
212
- "quantization_status": "compressed",
213
- "sparsity_config": {},
214
- "transform_config": {},
215
- "version": "0.12.3.a20251114"
216
- },
217
- "spatial_merge_size": 2,
218
- "text_config": {
219
- "attention_dropout": 0.0,
220
- "head_dim": 128,
221
- "hidden_act": "silu",
222
- "hidden_size": 5120,
223
- "initializer_range": 0.02,
224
- "intermediate_size": 32768,
225
- "max_position_embeddings": 393216,
226
- "model_type": "ministral3",
227
- "num_attention_heads": 32,
228
- "num_hidden_layers": 40,
229
- "num_key_value_heads": 8,
230
- "rms_norm_eps": 1e-05,
231
- "rope_parameters": {
232
- "beta_fast": 32.0,
233
- "beta_slow": 1.0,
234
- "factor": 48.0,
235
- "llama_4_scaling_beta": 0.1,
236
- "mscale": 1.0,
237
- "mscale_all_dim": 1.0,
238
- "original_max_position_embeddings": 8192,
239
- "rope_theta": 100000000.0,
240
- "rope_type": "yarn",
241
- "type": "yarn"
242
- },
243
- "sliding_window": null,
244
- "use_cache": true,
245
- "vocab_size": 131072
246
- },
247
- "transformers_version": "5.0.0.dev0",
248
- "vision_config": {
249
- "attention_dropout": 0.0,
250
- "head_dim": 64,
251
- "hidden_act": "silu",
252
- "hidden_size": 1024,
253
- "image_size": 1540,
254
- "initializer_range": 0.02,
255
- "intermediate_size": 4096,
256
- "model_type": "pixtral",
257
- "num_attention_heads": 16,
258
- "num_channels": 3,
259
- "num_hidden_layers": 24,
260
- "patch_size": 14,
261
- "rope_parameters": {
262
- "rope_theta": 10000.0,
263
- "rope_type": "default"
264
- }
265
- },
266
- "vision_feature_layer": -1
267
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
.ipynb_checkpoints/model.safetensors.index-checkpoint.json DELETED
The diff for this file is too large to render. See raw diff