ajaymin28 commited on
Commit
7d4de3f
1 Parent(s): 9fde4fd

Upload folder using huggingface_hub

Browse files
llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split023_all_mm_tune_olora256_512_llm/README.md ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+ ### Framework versions
7
+
8
+
9
+ - PEFT 0.4.0
llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split023_all_mm_tune_olora256_512_llm/adapter_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "lmms-lab/LLaVA-Video-7B-Qwen2",
4
+ "bias": "none",
5
+ "fan_in_fan_out": false,
6
+ "inference_mode": true,
7
+ "init_lora_weights": "olora",
8
+ "layers_pattern": null,
9
+ "layers_to_transform": null,
10
+ "lora_alpha": 512,
11
+ "lora_dropout": 0.05,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 256,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "gate_proj",
18
+ "o_proj",
19
+ "down_proj",
20
+ "up_proj",
21
+ "q_proj",
22
+ "k_proj",
23
+ "v_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split023_all_mm_tune_olora256_512_llm/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3bd98776562cf20502f8de947c0d1bb29d58b4040fd3c2afa894c0268d166f5a
3
+ size 1384057050
llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split023_all_mm_tune_olora256_512_llm/config.json ADDED
@@ -0,0 +1,221 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "lmms-lab/LLaVA-Video-7B-Qwen2",
3
+ "add_faster_video": false,
4
+ "add_time_instruction": true,
5
+ "architectures": [
6
+ "LlavaQwenForCausalLM"
7
+ ],
8
+ "attention_dropout": 0.0,
9
+ "bos_token_id": 151643,
10
+ "eos_token_id": 151645,
11
+ "faster_token_stride": 10,
12
+ "force_sample": true,
13
+ "hidden_act": "silu",
14
+ "hidden_size": 3584,
15
+ "ignore_index": -100,
16
+ "image_aspect_ratio": "anyres_max_9",
17
+ "image_crop_resolution": null,
18
+ "image_grid_pinpoints": [
19
+ [
20
+ 384,
21
+ 384
22
+ ],
23
+ [
24
+ 384,
25
+ 768
26
+ ],
27
+ [
28
+ 384,
29
+ 1152
30
+ ],
31
+ [
32
+ 384,
33
+ 1536
34
+ ],
35
+ [
36
+ 384,
37
+ 1920
38
+ ],
39
+ [
40
+ 384,
41
+ 2304
42
+ ],
43
+ [
44
+ 768,
45
+ 384
46
+ ],
47
+ [
48
+ 768,
49
+ 768
50
+ ],
51
+ [
52
+ 768,
53
+ 1152
54
+ ],
55
+ [
56
+ 768,
57
+ 1536
58
+ ],
59
+ [
60
+ 768,
61
+ 1920
62
+ ],
63
+ [
64
+ 768,
65
+ 2304
66
+ ],
67
+ [
68
+ 1152,
69
+ 384
70
+ ],
71
+ [
72
+ 1152,
73
+ 768
74
+ ],
75
+ [
76
+ 1152,
77
+ 1152
78
+ ],
79
+ [
80
+ 1152,
81
+ 1536
82
+ ],
83
+ [
84
+ 1152,
85
+ 1920
86
+ ],
87
+ [
88
+ 1152,
89
+ 2304
90
+ ],
91
+ [
92
+ 1536,
93
+ 384
94
+ ],
95
+ [
96
+ 1536,
97
+ 768
98
+ ],
99
+ [
100
+ 1536,
101
+ 1152
102
+ ],
103
+ [
104
+ 1536,
105
+ 1536
106
+ ],
107
+ [
108
+ 1536,
109
+ 1920
110
+ ],
111
+ [
112
+ 1536,
113
+ 2304
114
+ ],
115
+ [
116
+ 1920,
117
+ 384
118
+ ],
119
+ [
120
+ 1920,
121
+ 768
122
+ ],
123
+ [
124
+ 1920,
125
+ 1152
126
+ ],
127
+ [
128
+ 1920,
129
+ 1536
130
+ ],
131
+ [
132
+ 1920,
133
+ 1920
134
+ ],
135
+ [
136
+ 1920,
137
+ 2304
138
+ ],
139
+ [
140
+ 2304,
141
+ 384
142
+ ],
143
+ [
144
+ 2304,
145
+ 768
146
+ ],
147
+ [
148
+ 2304,
149
+ 1152
150
+ ],
151
+ [
152
+ 2304,
153
+ 1536
154
+ ],
155
+ [
156
+ 2304,
157
+ 1920
158
+ ],
159
+ [
160
+ 2304,
161
+ 2304
162
+ ]
163
+ ],
164
+ "image_split_resolution": null,
165
+ "image_token_index": 151646,
166
+ "initializer_range": 0.02,
167
+ "intermediate_size": 18944,
168
+ "max_position_embeddings": 32768,
169
+ "max_window_layers": 28,
170
+ "mm_hidden_size": 1152,
171
+ "mm_newline_position": "grid",
172
+ "mm_patch_merge_type": "spatial_unpad",
173
+ "mm_projector_lr": 2e-05,
174
+ "mm_projector_type": "mlp2x_gelu",
175
+ "mm_resampler_type": null,
176
+ "mm_spatial_pool_mode": "bilinear",
177
+ "mm_spatial_pool_stride": 2,
178
+ "mm_tunable_parts": "mm_vision_tower,mm_mlp_adapter,mm_language_model",
179
+ "mm_use_im_patch_token": false,
180
+ "mm_use_im_start_end": false,
181
+ "mm_vision_select_feature": "patch",
182
+ "mm_vision_select_layer": -2,
183
+ "mm_vision_tower": "google/siglip-so400m-patch14-384",
184
+ "mm_vision_tower_lr": null,
185
+ "model_type": "llava",
186
+ "num_attention_heads": 28,
187
+ "num_hidden_layers": 28,
188
+ "num_key_value_heads": 4,
189
+ "pos_skipping_range": 4096,
190
+ "projector_hidden_act": "gelu",
191
+ "rms_norm_eps": 1e-06,
192
+ "rope_scaling": null,
193
+ "rope_theta": 1000000.0,
194
+ "sliding_window": 131072,
195
+ "text_config": {
196
+ "model_type": "llama"
197
+ },
198
+ "tie_word_embeddings": false,
199
+ "tokenizer_model_max_length": 32768,
200
+ "tokenizer_padding_side": "right",
201
+ "torch_dtype": "bfloat16",
202
+ "transformers_version": "4.40.0.dev0",
203
+ "use_cache": true,
204
+ "use_mm_proj": true,
205
+ "use_pos_skipping": false,
206
+ "use_sliding_window": false,
207
+ "vision_config": {
208
+ "hidden_size": 1024,
209
+ "image_size": 336,
210
+ "intermediate_size": 4096,
211
+ "model_type": "clip_vision_model",
212
+ "num_attention_heads": 16,
213
+ "num_hidden_layers": 24,
214
+ "patch_size": 14,
215
+ "projection_dim": 768,
216
+ "vocab_size": 32000
217
+ },
218
+ "vision_feature_layer": -2,
219
+ "vision_feature_select_strategy": "default",
220
+ "vision_tower_pretrained": null
221
+ }
llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split023_all_mm_tune_olora256_512_llm/generation_config.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 151643,
3
+ "do_sample": true,
4
+ "eos_token_id": [
5
+ 151645,
6
+ 151643
7
+ ],
8
+ "pad_token_id": 151643,
9
+ "repetition_penalty": 1.05,
10
+ "temperature": 0.7,
11
+ "top_k": 20,
12
+ "top_p": 0.8,
13
+ "transformers_version": "4.40.0.dev0"
14
+ }
llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split023_all_mm_tune_olora256_512_llm/non_lora_trainables.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:30ae9c70d971e0301f679ca59b8893a810cd89ff10ff9bbb12753d7df9ce1b06
3
+ size 33964208
llavaqwen2-google_siglip-so400m-patch14-384-Qwen_Qwen2-7B-Instruct-AG_v5_3_split023_all_mm_tune_olora256_512_llm/trainer_state.json ADDED
The diff for this file is too large to render. See raw diff