Hugging Face's logo

nvidia
/
Flora-TempSeg

PEFT
Safetensors
lora
qwen3.5
video
temporal-segmentation
data-engine
Model card Files
xet
 Community
6
Omar L commited on
Commit
143ed82
·
verified ·
1 Parent(s): b2d4169

Stage final Flora-TempSeg weights

Browse files

Source: flora-share/Flora-TempSeg-Release@fc92075dc0b6b2e73d4f48f55a6fc1c034e8ed16

Files changed (9) hide show
  1. .gitattributes +1 -0
  2. MODEL_PROVENANCE.json +13 -21
  3. SOURCE_MANIFEST.sha256 +8 -2
  4. adapter_config.json +251 -41
  5. adapter_model.safetensors +2 -2
  6. chat_template.jinja +48 -0
  7. processor_config.json +60 -0
  8. tokenizer.json +3 -0
  9. tokenizer_config.json +35 -0
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
MODEL_PROVENANCE.json CHANGED
@@ -1,24 +1,16 @@
1
  {
2
- "artifact_kind": "peft_lora_adapter",
 
 
3
  "base_model": "Qwen/Qwen3.5-9B",
4
- "base_model_revision": "c202236235762e1c871ad0ccb60c8ee5ba337b9a",
5
- "checkpoint": "step-5114",
6
- "created_utc": "2026-08-25T08:19:35Z",
7
- "inference_stage": "tempseg",
8
- "schema_version": "nvember_data_engine_model_provenance_v1",
9
- "source_full_sha256_verified": true,
10
- "source_mutated": false,
11
- "training_lineage": "mecka_v3_redelivered_t2",
12
- "packaging_adjustments": [
13
- {
14
- "file": "adapter_config.json",
15
- "change": {
16
- "base_model_name_or_path": "Qwen/Qwen3.5-9B",
17
- "revision": "c202236235762e1c871ad0ccb60c8ee5ba337b9a"
18
- },
19
- "original_sha256": "bd1f51e21c38512f0e4a9b63a502c1282540e9c53dc5417951f36957cc163b61",
20
- "packaged_sha256": "ecb617b821c2c1dea8ad6bacc7ed312d7dadb90c73b5cfe4ec32d33fc0ebb4ee",
21
- "weights_changed": false
22
- }
23
- ]
24
  }
 
1
  {
2
+ "schema_version": 2,
3
+ "model": "Flora-segmenter-9B",
4
+ "selected_step": 1300,
5
  "base_model": "Qwen/Qwen3.5-9B",
6
+ "base_revision": "c202236235762e1c871ad0ccb60c8ee5ba337b9a",
7
+ "format": "PEFT LoRA",
8
+ "adapter_sha256": "1d80fe544e4cd1a7ff6b0ea4be71e988972408dab81d4c8af72e3a628739d6c3",
9
+ "adapter_tensor_count": 496,
10
+ "adapter_dtype": "F32",
11
+ "lora_rank": 32,
12
+ "lora_alpha": 64,
13
+ "lora_scope": "language-model attention and MLP projections",
14
+ "source_weights_modified": false,
15
+ "adapter_config_public_base_binding_only": true
 
 
 
 
 
 
 
 
 
 
16
  }
SOURCE_MANIFEST.sha256 CHANGED
@@ -1,2 +1,8 @@
1
- ecb617b821c2c1dea8ad6bacc7ed312d7dadb90c73b5cfe4ec32d33fc0ebb4ee adapter_config.json
2
- 13a0eaa6f45bc1cc39c989c7b8d495b64aedc34839f0860d0854b01afe92ac34 adapter_model.safetensors
 
 
 
 
 
 
 
1
+ 837c0aa8748126d2193799e7bec51d6550ca4ba07767591e66dcb370c3115e44 MODEL_PROVENANCE.json
2
+ 6c051cfaeac3bbce138ba1b95ff1e21c7f2812356c472e9a11cf4b9467099680 README.md
3
+ ac4d8abd2a14045b22151354a7493049d651ec10c0b019fbe4927805e45c4b81 adapter_config.json
4
+ 1d80fe544e4cd1a7ff6b0ea4be71e988972408dab81d4c8af72e3a628739d6c3 adapter_model.safetensors
5
+ d28ca68768ac2ce0ad42eaaf54d5bac4f8a298f7d017f5a04f3866094c65a839 chat_template.jinja
6
+ efcc4bbf42f0e1c98d1804dc6484590d2ac3d1ad2575f67b4a59b5706ea8fa86 processor_config.json
7
+ 06b9509352d2af50381ab2247e083b80d32d5c0aba91c272ca9ff729b6a0e523 tokenizer.json
8
+ b4c4e0a37983a58c329c8c4e25d59f3305fd344815f4ecb8c66a99323a254cf8 tokenizer_config.json
adapter_config.json CHANGED
@@ -1,51 +1,261 @@
1
  {
2
- "alora_invocation_tokens": null,
3
- "alpha_pattern": {},
4
- "arrow_config": null,
5
- "auto_mapping": null,
6
  "base_model_name_or_path": "Qwen/Qwen3.5-9B",
7
  "bias": "none",
8
- "corda_config": null,
9
- "ensure_weight_tying": false,
10
- "eva_config": null,
11
- "exclude_modules": null,
12
- "fan_in_fan_out": false,
13
- "inference_mode": true,
14
- "init_lora_weights": true,
15
- "layer_replication": null,
16
- "layers_pattern": null,
17
- "layers_to_transform": null,
18
- "loftq_config": {},
19
  "lora_alpha": 64,
20
- "lora_bias": false,
21
- "lora_dropout": 0.0,
22
- "megatron_config": null,
23
- "megatron_core": "megatron.core",
24
- "modules_to_save": null,
25
  "peft_type": "LORA",
26
- "peft_version": "0.18.1",
27
- "qalora_group_size": 16,
28
  "r": 32,
29
- "rank_pattern": {},
30
- "revision": "c202236235762e1c871ad0ccb60c8ee5ba337b9a",
31
  "target_modules": [
32
- "up_proj",
33
- "k_proj",
34
- "in_proj_b",
35
- "out_proj",
36
- "v_proj",
37
- "in_proj_z",
38
- "q_proj",
39
- "o_proj",
40
- "in_proj_a",
41
- "in_proj_qkv",
42
- "gate_proj",
43
- "down_proj"
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
44
  ],
45
- "target_parameters": null,
46
  "task_type": "CAUSAL_LM",
47
- "trainable_token_indices": null,
48
  "use_dora": false,
49
- "use_qalora": false,
50
- "use_rslora": false
51
- }
 
1
  {
 
 
 
 
2
  "base_model_name_or_path": "Qwen/Qwen3.5-9B",
3
  "bias": "none",
 
 
 
 
 
 
 
 
 
 
 
4
  "lora_alpha": 64,
 
 
 
 
 
5
  "peft_type": "LORA",
 
 
6
  "r": 32,
 
 
7
  "target_modules": [
8
+ "model.language_model.layers.0.linear_attn.in_proj_a",
9
+ "model.language_model.layers.0.linear_attn.in_proj_b",
10
+ "model.language_model.layers.0.linear_attn.in_proj_qkv",
11
+ "model.language_model.layers.0.linear_attn.in_proj_z",
12
+ "model.language_model.layers.0.linear_attn.out_proj",
13
+ "model.language_model.layers.0.mlp.down_proj",
14
+ "model.language_model.layers.0.mlp.gate_proj",
15
+ "model.language_model.layers.0.mlp.up_proj",
16
+ "model.language_model.layers.1.linear_attn.in_proj_a",
17
+ "model.language_model.layers.1.linear_attn.in_proj_b",
18
+ "model.language_model.layers.1.linear_attn.in_proj_qkv",
19
+ "model.language_model.layers.1.linear_attn.in_proj_z",
20
+ "model.language_model.layers.1.linear_attn.out_proj",
21
+ "model.language_model.layers.1.mlp.down_proj",
22
+ "model.language_model.layers.1.mlp.gate_proj",
23
+ "model.language_model.layers.1.mlp.up_proj",
24
+ "model.language_model.layers.10.linear_attn.in_proj_a",
25
+ "model.language_model.layers.10.linear_attn.in_proj_b",
26
+ "model.language_model.layers.10.linear_attn.in_proj_qkv",
27
+ "model.language_model.layers.10.linear_attn.in_proj_z",
28
+ "model.language_model.layers.10.linear_attn.out_proj",
29
+ "model.language_model.layers.10.mlp.down_proj",
30
+ "model.language_model.layers.10.mlp.gate_proj",
31
+ "model.language_model.layers.10.mlp.up_proj",
32
+ "model.language_model.layers.11.mlp.down_proj",
33
+ "model.language_model.layers.11.mlp.gate_proj",
34
+ "model.language_model.layers.11.mlp.up_proj",
35
+ "model.language_model.layers.11.self_attn.k_proj",
36
+ "model.language_model.layers.11.self_attn.o_proj",
37
+ "model.language_model.layers.11.self_attn.q_proj",
38
+ "model.language_model.layers.11.self_attn.v_proj",
39
+ "model.language_model.layers.12.linear_attn.in_proj_a",
40
+ "model.language_model.layers.12.linear_attn.in_proj_b",
41
+ "model.language_model.layers.12.linear_attn.in_proj_qkv",
42
+ "model.language_model.layers.12.linear_attn.in_proj_z",
43
+ "model.language_model.layers.12.linear_attn.out_proj",
44
+ "model.language_model.layers.12.mlp.down_proj",
45
+ "model.language_model.layers.12.mlp.gate_proj",
46
+ "model.language_model.layers.12.mlp.up_proj",
47
+ "model.language_model.layers.13.linear_attn.in_proj_a",
48
+ "model.language_model.layers.13.linear_attn.in_proj_b",
49
+ "model.language_model.layers.13.linear_attn.in_proj_qkv",
50
+ "model.language_model.layers.13.linear_attn.in_proj_z",
51
+ "model.language_model.layers.13.linear_attn.out_proj",
52
+ "model.language_model.layers.13.mlp.down_proj",
53
+ "model.language_model.layers.13.mlp.gate_proj",
54
+ "model.language_model.layers.13.mlp.up_proj",
55
+ "model.language_model.layers.14.linear_attn.in_proj_a",
56
+ "model.language_model.layers.14.linear_attn.in_proj_b",
57
+ "model.language_model.layers.14.linear_attn.in_proj_qkv",
58
+ "model.language_model.layers.14.linear_attn.in_proj_z",
59
+ "model.language_model.layers.14.linear_attn.out_proj",
60
+ "model.language_model.layers.14.mlp.down_proj",
61
+ "model.language_model.layers.14.mlp.gate_proj",
62
+ "model.language_model.layers.14.mlp.up_proj",
63
+ "model.language_model.layers.15.mlp.down_proj",
64
+ "model.language_model.layers.15.mlp.gate_proj",
65
+ "model.language_model.layers.15.mlp.up_proj",
66
+ "model.language_model.layers.15.self_attn.k_proj",
67
+ "model.language_model.layers.15.self_attn.o_proj",
68
+ "model.language_model.layers.15.self_attn.q_proj",
69
+ "model.language_model.layers.15.self_attn.v_proj",
70
+ "model.language_model.layers.16.linear_attn.in_proj_a",
71
+ "model.language_model.layers.16.linear_attn.in_proj_b",
72
+ "model.language_model.layers.16.linear_attn.in_proj_qkv",
73
+ "model.language_model.layers.16.linear_attn.in_proj_z",
74
+ "model.language_model.layers.16.linear_attn.out_proj",
75
+ "model.language_model.layers.16.mlp.down_proj",
76
+ "model.language_model.layers.16.mlp.gate_proj",
77
+ "model.language_model.layers.16.mlp.up_proj",
78
+ "model.language_model.layers.17.linear_attn.in_proj_a",
79
+ "model.language_model.layers.17.linear_attn.in_proj_b",
80
+ "model.language_model.layers.17.linear_attn.in_proj_qkv",
81
+ "model.language_model.layers.17.linear_attn.in_proj_z",
82
+ "model.language_model.layers.17.linear_attn.out_proj",
83
+ "model.language_model.layers.17.mlp.down_proj",
84
+ "model.language_model.layers.17.mlp.gate_proj",
85
+ "model.language_model.layers.17.mlp.up_proj",
86
+ "model.language_model.layers.18.linear_attn.in_proj_a",
87
+ "model.language_model.layers.18.linear_attn.in_proj_b",
88
+ "model.language_model.layers.18.linear_attn.in_proj_qkv",
89
+ "model.language_model.layers.18.linear_attn.in_proj_z",
90
+ "model.language_model.layers.18.linear_attn.out_proj",
91
+ "model.language_model.layers.18.mlp.down_proj",
92
+ "model.language_model.layers.18.mlp.gate_proj",
93
+ "model.language_model.layers.18.mlp.up_proj",
94
+ "model.language_model.layers.19.mlp.down_proj",
95
+ "model.language_model.layers.19.mlp.gate_proj",
96
+ "model.language_model.layers.19.mlp.up_proj",
97
+ "model.language_model.layers.19.self_attn.k_proj",
98
+ "model.language_model.layers.19.self_attn.o_proj",
99
+ "model.language_model.layers.19.self_attn.q_proj",
100
+ "model.language_model.layers.19.self_attn.v_proj",
101
+ "model.language_model.layers.2.linear_attn.in_proj_a",
102
+ "model.language_model.layers.2.linear_attn.in_proj_b",
103
+ "model.language_model.layers.2.linear_attn.in_proj_qkv",
104
+ "model.language_model.layers.2.linear_attn.in_proj_z",
105
+ "model.language_model.layers.2.linear_attn.out_proj",
106
+ "model.language_model.layers.2.mlp.down_proj",
107
+ "model.language_model.layers.2.mlp.gate_proj",
108
+ "model.language_model.layers.2.mlp.up_proj",
109
+ "model.language_model.layers.20.linear_attn.in_proj_a",
110
+ "model.language_model.layers.20.linear_attn.in_proj_b",
111
+ "model.language_model.layers.20.linear_attn.in_proj_qkv",
112
+ "model.language_model.layers.20.linear_attn.in_proj_z",
113
+ "model.language_model.layers.20.linear_attn.out_proj",
114
+ "model.language_model.layers.20.mlp.down_proj",
115
+ "model.language_model.layers.20.mlp.gate_proj",
116
+ "model.language_model.layers.20.mlp.up_proj",
117
+ "model.language_model.layers.21.linear_attn.in_proj_a",
118
+ "model.language_model.layers.21.linear_attn.in_proj_b",
119
+ "model.language_model.layers.21.linear_attn.in_proj_qkv",
120
+ "model.language_model.layers.21.linear_attn.in_proj_z",
121
+ "model.language_model.layers.21.linear_attn.out_proj",
122
+ "model.language_model.layers.21.mlp.down_proj",
123
+ "model.language_model.layers.21.mlp.gate_proj",
124
+ "model.language_model.layers.21.mlp.up_proj",
125
+ "model.language_model.layers.22.linear_attn.in_proj_a",
126
+ "model.language_model.layers.22.linear_attn.in_proj_b",
127
+ "model.language_model.layers.22.linear_attn.in_proj_qkv",
128
+ "model.language_model.layers.22.linear_attn.in_proj_z",
129
+ "model.language_model.layers.22.linear_attn.out_proj",
130
+ "model.language_model.layers.22.mlp.down_proj",
131
+ "model.language_model.layers.22.mlp.gate_proj",
132
+ "model.language_model.layers.22.mlp.up_proj",
133
+ "model.language_model.layers.23.mlp.down_proj",
134
+ "model.language_model.layers.23.mlp.gate_proj",
135
+ "model.language_model.layers.23.mlp.up_proj",
136
+ "model.language_model.layers.23.self_attn.k_proj",
137
+ "model.language_model.layers.23.self_attn.o_proj",
138
+ "model.language_model.layers.23.self_attn.q_proj",
139
+ "model.language_model.layers.23.self_attn.v_proj",
140
+ "model.language_model.layers.24.linear_attn.in_proj_a",
141
+ "model.language_model.layers.24.linear_attn.in_proj_b",
142
+ "model.language_model.layers.24.linear_attn.in_proj_qkv",
143
+ "model.language_model.layers.24.linear_attn.in_proj_z",
144
+ "model.language_model.layers.24.linear_attn.out_proj",
145
+ "model.language_model.layers.24.mlp.down_proj",
146
+ "model.language_model.layers.24.mlp.gate_proj",
147
+ "model.language_model.layers.24.mlp.up_proj",
148
+ "model.language_model.layers.25.linear_attn.in_proj_a",
149
+ "model.language_model.layers.25.linear_attn.in_proj_b",
150
+ "model.language_model.layers.25.linear_attn.in_proj_qkv",
151
+ "model.language_model.layers.25.linear_attn.in_proj_z",
152
+ "model.language_model.layers.25.linear_attn.out_proj",
153
+ "model.language_model.layers.25.mlp.down_proj",
154
+ "model.language_model.layers.25.mlp.gate_proj",
155
+ "model.language_model.layers.25.mlp.up_proj",
156
+ "model.language_model.layers.26.linear_attn.in_proj_a",
157
+ "model.language_model.layers.26.linear_attn.in_proj_b",
158
+ "model.language_model.layers.26.linear_attn.in_proj_qkv",
159
+ "model.language_model.layers.26.linear_attn.in_proj_z",
160
+ "model.language_model.layers.26.linear_attn.out_proj",
161
+ "model.language_model.layers.26.mlp.down_proj",
162
+ "model.language_model.layers.26.mlp.gate_proj",
163
+ "model.language_model.layers.26.mlp.up_proj",
164
+ "model.language_model.layers.27.mlp.down_proj",
165
+ "model.language_model.layers.27.mlp.gate_proj",
166
+ "model.language_model.layers.27.mlp.up_proj",
167
+ "model.language_model.layers.27.self_attn.k_proj",
168
+ "model.language_model.layers.27.self_attn.o_proj",
169
+ "model.language_model.layers.27.self_attn.q_proj",
170
+ "model.language_model.layers.27.self_attn.v_proj",
171
+ "model.language_model.layers.28.linear_attn.in_proj_a",
172
+ "model.language_model.layers.28.linear_attn.in_proj_b",
173
+ "model.language_model.layers.28.linear_attn.in_proj_qkv",
174
+ "model.language_model.layers.28.linear_attn.in_proj_z",
175
+ "model.language_model.layers.28.linear_attn.out_proj",
176
+ "model.language_model.layers.28.mlp.down_proj",
177
+ "model.language_model.layers.28.mlp.gate_proj",
178
+ "model.language_model.layers.28.mlp.up_proj",
179
+ "model.language_model.layers.29.linear_attn.in_proj_a",
180
+ "model.language_model.layers.29.linear_attn.in_proj_b",
181
+ "model.language_model.layers.29.linear_attn.in_proj_qkv",
182
+ "model.language_model.layers.29.linear_attn.in_proj_z",
183
+ "model.language_model.layers.29.linear_attn.out_proj",
184
+ "model.language_model.layers.29.mlp.down_proj",
185
+ "model.language_model.layers.29.mlp.gate_proj",
186
+ "model.language_model.layers.29.mlp.up_proj",
187
+ "model.language_model.layers.3.mlp.down_proj",
188
+ "model.language_model.layers.3.mlp.gate_proj",
189
+ "model.language_model.layers.3.mlp.up_proj",
190
+ "model.language_model.layers.3.self_attn.k_proj",
191
+ "model.language_model.layers.3.self_attn.o_proj",
192
+ "model.language_model.layers.3.self_attn.q_proj",
193
+ "model.language_model.layers.3.self_attn.v_proj",
194
+ "model.language_model.layers.30.linear_attn.in_proj_a",
195
+ "model.language_model.layers.30.linear_attn.in_proj_b",
196
+ "model.language_model.layers.30.linear_attn.in_proj_qkv",
197
+ "model.language_model.layers.30.linear_attn.in_proj_z",
198
+ "model.language_model.layers.30.linear_attn.out_proj",
199
+ "model.language_model.layers.30.mlp.down_proj",
200
+ "model.language_model.layers.30.mlp.gate_proj",
201
+ "model.language_model.layers.30.mlp.up_proj",
202
+ "model.language_model.layers.31.mlp.down_proj",
203
+ "model.language_model.layers.31.mlp.gate_proj",
204
+ "model.language_model.layers.31.mlp.up_proj",
205
+ "model.language_model.layers.31.self_attn.k_proj",
206
+ "model.language_model.layers.31.self_attn.o_proj",
207
+ "model.language_model.layers.31.self_attn.q_proj",
208
+ "model.language_model.layers.31.self_attn.v_proj",
209
+ "model.language_model.layers.4.linear_attn.in_proj_a",
210
+ "model.language_model.layers.4.linear_attn.in_proj_b",
211
+ "model.language_model.layers.4.linear_attn.in_proj_qkv",
212
+ "model.language_model.layers.4.linear_attn.in_proj_z",
213
+ "model.language_model.layers.4.linear_attn.out_proj",
214
+ "model.language_model.layers.4.mlp.down_proj",
215
+ "model.language_model.layers.4.mlp.gate_proj",
216
+ "model.language_model.layers.4.mlp.up_proj",
217
+ "model.language_model.layers.5.linear_attn.in_proj_a",
218
+ "model.language_model.layers.5.linear_attn.in_proj_b",
219
+ "model.language_model.layers.5.linear_attn.in_proj_qkv",
220
+ "model.language_model.layers.5.linear_attn.in_proj_z",
221
+ "model.language_model.layers.5.linear_attn.out_proj",
222
+ "model.language_model.layers.5.mlp.down_proj",
223
+ "model.language_model.layers.5.mlp.gate_proj",
224
+ "model.language_model.layers.5.mlp.up_proj",
225
+ "model.language_model.layers.6.linear_attn.in_proj_a",
226
+ "model.language_model.layers.6.linear_attn.in_proj_b",
227
+ "model.language_model.layers.6.linear_attn.in_proj_qkv",
228
+ "model.language_model.layers.6.linear_attn.in_proj_z",
229
+ "model.language_model.layers.6.linear_attn.out_proj",
230
+ "model.language_model.layers.6.mlp.down_proj",
231
+ "model.language_model.layers.6.mlp.gate_proj",
232
+ "model.language_model.layers.6.mlp.up_proj",
233
+ "model.language_model.layers.7.mlp.down_proj",
234
+ "model.language_model.layers.7.mlp.gate_proj",
235
+ "model.language_model.layers.7.mlp.up_proj",
236
+ "model.language_model.layers.7.self_attn.k_proj",
237
+ "model.language_model.layers.7.self_attn.o_proj",
238
+ "model.language_model.layers.7.self_attn.q_proj",
239
+ "model.language_model.layers.7.self_attn.v_proj",
240
+ "model.language_model.layers.8.linear_attn.in_proj_a",
241
+ "model.language_model.layers.8.linear_attn.in_proj_b",
242
+ "model.language_model.layers.8.linear_attn.in_proj_qkv",
243
+ "model.language_model.layers.8.linear_attn.in_proj_z",
244
+ "model.language_model.layers.8.linear_attn.out_proj",
245
+ "model.language_model.layers.8.mlp.down_proj",
246
+ "model.language_model.layers.8.mlp.gate_proj",
247
+ "model.language_model.layers.8.mlp.up_proj",
248
+ "model.language_model.layers.9.linear_attn.in_proj_a",
249
+ "model.language_model.layers.9.linear_attn.in_proj_b",
250
+ "model.language_model.layers.9.linear_attn.in_proj_qkv",
251
+ "model.language_model.layers.9.linear_attn.in_proj_z",
252
+ "model.language_model.layers.9.linear_attn.out_proj",
253
+ "model.language_model.layers.9.mlp.down_proj",
254
+ "model.language_model.layers.9.mlp.gate_proj",
255
+ "model.language_model.layers.9.mlp.up_proj"
256
  ],
 
257
  "task_type": "CAUSAL_LM",
 
258
  "use_dora": false,
259
+ "revision": "c202236235762e1c871ad0ccb60c8ee5ba337b9a",
260
+ "inference_mode": true
261
+ }
adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:13a0eaa6f45bc1cc39c989c7b8d495b64aedc34839f0860d0854b01afe92ac34
3
- size 173189008
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1d80fe544e4cd1a7ff6b0ea4be71e988972408dab81d4c8af72e3a628739d6c3
3
+ size 346302152
chat_template.jinja ADDED
@@ -0,0 +1,48 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {%- macro render_content(content, is_system_content=false) %}
2
+ {%- if content is string %}
3
+ {{- content }}
4
+ {%- elif content is iterable and content is not mapping %}
5
+ {%- for item in content %}
6
+ {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}
7
+ {%- if is_system_content %}
8
+ {{- raise_exception('System message cannot contain images.') }}
9
+ {%- endif %}
10
+ {{- '<|vision_start|><|image_pad|><|vision_end|>' }}
11
+ {%- elif 'video' in item or item.type == 'video' %}
12
+ {%- if is_system_content %}
13
+ {{- raise_exception('System message cannot contain videos.') }}
14
+ {%- endif %}
15
+ {{- '<|vision_start|><|video_pad|><|vision_end|>' }}
16
+ {%- elif 'text' in item %}
17
+ {{- item.text }}
18
+ {%- else %}
19
+ {{- raise_exception('Unexpected item type in content.') }}
20
+ {%- endif %}
21
+ {%- endfor %}
22
+ {%- elif content is none or content is undefined %}
23
+ {{- '' }}
24
+ {%- else %}
25
+ {{- raise_exception('Unexpected content type.') }}
26
+ {%- endif %}
27
+ {%- endmacro %}
28
+ {%- if not messages %}
29
+ {{- raise_exception('No messages provided.') }}
30
+ {%- endif %}
31
+ {%- if tools %}
32
+ {{- raise_exception('This chat template does not support tools.') }}
33
+ {%- endif %}
34
+ {%- for message in messages %}
35
+ {%- if message.role == 'system' %}
36
+ {%- if not loop.first %}
37
+ {{- raise_exception('System message must be first.') }}
38
+ {%- endif %}
39
+ {{- '<|im_start|>system\n' + render_content(message.content, true) + '<|im_end|>\n' }}
40
+ {%- elif message.role == 'user' or message.role == 'assistant' %}
41
+ {{- '<|im_start|>' + message.role + '\n' + render_content(message.content) + '<|im_end|>\n' }}
42
+ {%- else %}
43
+ {{- raise_exception('Only system/user/assistant roles are supported.') }}
44
+ {%- endif %}
45
+ {%- endfor %}
46
+ {%- if add_generation_prompt %}
47
+ {{- '<|im_start|>assistant\n' }}
48
+ {%- endif %}
processor_config.json ADDED
@@ -0,0 +1,60 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "image_processor": {
3
+ "do_convert_rgb": true,
4
+ "do_normalize": true,
5
+ "do_rescale": true,
6
+ "do_resize": true,
7
+ "image_mean": [
8
+ 0.5,
9
+ 0.5,
10
+ 0.5
11
+ ],
12
+ "image_processor_type": "Qwen2VLImageProcessor",
13
+ "image_std": [
14
+ 0.5,
15
+ 0.5,
16
+ 0.5
17
+ ],
18
+ "merge_size": 2,
19
+ "patch_size": 16,
20
+ "resample": 3,
21
+ "rescale_factor": 0.00392156862745098,
22
+ "size": {
23
+ "longest_edge": 589824,
24
+ "shortest_edge": 1024
25
+ },
26
+ "temporal_patch_size": 2
27
+ },
28
+ "processor_class": "Qwen3VLProcessor",
29
+ "video_processor": {
30
+ "do_convert_rgb": true,
31
+ "do_normalize": true,
32
+ "do_rescale": true,
33
+ "do_resize": true,
34
+ "do_sample_frames": true,
35
+ "fps": 4,
36
+ "image_mean": [
37
+ 0.5,
38
+ 0.5,
39
+ 0.5
40
+ ],
41
+ "image_std": [
42
+ 0.5,
43
+ 0.5,
44
+ 0.5
45
+ ],
46
+ "max_frames": 128,
47
+ "merge_size": 2,
48
+ "min_frames": 4,
49
+ "patch_size": 16,
50
+ "resample": 3,
51
+ "rescale_factor": 0.00392156862745098,
52
+ "return_metadata": false,
53
+ "size": {
54
+ "longest_edge": 25165824,
55
+ "shortest_edge": 4096
56
+ },
57
+ "temporal_patch_size": 2,
58
+ "video_processor_type": "Qwen3VLVideoProcessor"
59
+ }
60
+ }
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:06b9509352d2af50381ab2247e083b80d32d5c0aba91c272ca9ff729b6a0e523
3
+ size 19989325
tokenizer_config.json ADDED
@@ -0,0 +1,35 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "audio_bos_token": "<|audio_start|>",
4
+ "audio_eos_token": "<|audio_end|>",
5
+ "audio_token": "<|audio_pad|>",
6
+ "backend": "tokenizers",
7
+ "bos_token": null,
8
+ "clean_up_tokenization_spaces": false,
9
+ "eos_token": "<|im_end|>",
10
+ "errors": "replace",
11
+ "image_token": "<|image_pad|>",
12
+ "is_local": true,
13
+ "local_files_only": true,
14
+ "max_pixels": 589824,
15
+ "min_pixels": 1024,
16
+ "model_max_length": 262144,
17
+ "model_specific_special_tokens": {
18
+ "audio_bos_token": "<|audio_start|>",
19
+ "audio_eos_token": "<|audio_end|>",
20
+ "audio_token": "<|audio_pad|>",
21
+ "image_token": "<|image_pad|>",
22
+ "video_token": "<|video_pad|>",
23
+ "vision_bos_token": "<|vision_start|>",
24
+ "vision_eos_token": "<|vision_end|>"
25
+ },
26
+ "pad_token": "<|endoftext|>",
27
+ "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
28
+ "processor_class": "Qwen3VLProcessor",
29
+ "split_special_tokens": false,
30
+ "tokenizer_class": "Qwen2Tokenizer",
31
+ "unk_token": null,
32
+ "video_token": "<|video_pad|>",
33
+ "vision_bos_token": "<|vision_start|>",
34
+ "vision_eos_token": "<|vision_end|>"
35
+ }