Hugging Face's logo

AlanaAI
/
AlanaVLM

English
vlm
egocentric
embodied ai
Model card Files Community
claudiogreco commited on
Commit
11b6f93
•
1 Parent(s): f8856e0

Added checkpoints.

Browse files
Files changed (25) hide show
  1. alana-chatunivi-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/README.md +9 -0
  2. alana-chatunivi-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/adapter_config.json +27 -0
  3. alana-chatunivi-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/adapter_model.bin +3 -0
  4. alana-chatunivi-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/config.json +49 -0
  5. alana-chatunivi-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/non_lora_trainables.bin +3 -0
  6. alana-chatunivi-vqa-gemini-egoclip-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/README.md +9 -0
  7. alana-chatunivi-vqa-gemini-egoclip-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/adapter_config.json +27 -0
  8. alana-chatunivi-vqa-gemini-egoclip-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/adapter_model.bin +3 -0
  9. alana-chatunivi-vqa-gemini-egoclip-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/config.json +49 -0
  10. alana-chatunivi-vqa-gemini-egoclip-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/non_lora_trainables.bin +3 -0
  11. alana-chatunivi-vqa-gemini-egoclip-hm3d-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/README.md +9 -0
  12. alana-chatunivi-vqa-gemini-egoclip-hm3d-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/adapter_config.json +27 -0
  13. alana-chatunivi-vqa-gemini-egoclip-hm3d-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/adapter_model.bin +3 -0
  14. alana-chatunivi-vqa-gemini-egoclip-hm3d-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/config.json +49 -0
  15. alana-chatunivi-vqa-gemini-egoclip-hm3d-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/non_lora_trainables.bin +3 -0
  16. alana-chatunivi-vqa-gemini-vsr-egoclip-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/README.md +9 -0
  17. alana-chatunivi-vqa-gemini-vsr-egoclip-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/adapter_config.json +27 -0
  18. alana-chatunivi-vqa-gemini-vsr-egoclip-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/adapter_model.bin +3 -0
  19. alana-chatunivi-vqa-gemini-vsr-egoclip-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/config.json +49 -0
  20. alana-chatunivi-vqa-gemini-vsr-egoclip-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/non_lora_trainables.bin +3 -0
  21. alana-chatunivi-vqa-gemini-vsr-egoclip-hm3d-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/README.md +9 -0
  22. alana-chatunivi-vqa-gemini-vsr-egoclip-hm3d-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/adapter_config.json +27 -0
  23. alana-chatunivi-vqa-gemini-vsr-egoclip-hm3d-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/adapter_model.bin +3 -0
  24. alana-chatunivi-vqa-gemini-vsr-egoclip-hm3d-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/config.json +49 -0
  25. alana-chatunivi-vqa-gemini-vsr-egoclip-hm3d-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/non_lora_trainables.bin +3 -0
alana-chatunivi-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/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
alana-chatunivi-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/adapter_config.json ADDED
@@ -0,0 +1,27 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "Chat-UniVi/Chat-UniVi",
4
+ "bias": "none",
5
+ "fan_in_fan_out": false,
6
+ "inference_mode": true,
7
+ "init_lora_weights": true,
8
+ "layers_pattern": null,
9
+ "layers_to_transform": null,
10
+ "lora_alpha": 128,
11
+ "lora_dropout": 0.05,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 64,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "down_proj",
18
+ "gate_proj",
19
+ "o_proj",
20
+ "up_proj",
21
+ "q_proj",
22
+ "k_proj",
23
+ "mm_projector",
24
+ "v_proj"
25
+ ],
26
+ "task_type": "CAUSAL_LM"
27
+ }
alana-chatunivi-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:97bccecfe25460a2541551c4943261099f3a4f58591e973a59caf3fe583bc619
3
+ size 320627474
alana-chatunivi-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/config.json ADDED
@@ -0,0 +1,49 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "Chat-UniVi/Chat-UniVi",
3
+ "aarchitectures": "LlavaLlamaForCausalLM",
4
+ "architectures": [
5
+ "ChatUniViLlamaForCausalLM"
6
+ ],
7
+ "bos_token_id": 1,
8
+ "config": {
9
+ "freeze": false,
10
+ "mm_tune": true,
11
+ "spatial_cluster_rate0": 64,
12
+ "spatial_cluster_rate1": 32,
13
+ "spatial_cluster_rate2": 16,
14
+ "temporal_cluster_rate": 0.0625,
15
+ "use_cluster": true,
16
+ "vision_tune": false
17
+ },
18
+ "eos_token_id": 2,
19
+ "freeze_mm_mlp_adapter": false,
20
+ "hidden_act": "silu",
21
+ "hidden_size": 4096,
22
+ "image_aspect_ratio": "square",
23
+ "image_grid_pinpoints": null,
24
+ "initializer_range": 0.02,
25
+ "intermediate_size": 11008,
26
+ "max_position_embeddings": 4096,
27
+ "mm_hidden_size": 1024,
28
+ "mm_use_box_start_end": false,
29
+ "mm_use_im_patch_token": false,
30
+ "mm_use_im_start_end": false,
31
+ "mm_vision_select_feature": "patch",
32
+ "mm_vision_select_layer": -2,
33
+ "mm_vision_tower": "openai/clip-vit-large-patch14",
34
+ "model_type": "ChatUniVi",
35
+ "num_attention_heads": 32,
36
+ "num_hidden_layers": 32,
37
+ "num_key_value_heads": 32,
38
+ "pad_token_id": 0,
39
+ "pretraining_tp": 1,
40
+ "rms_norm_eps": 1e-05,
41
+ "rope_scaling": null,
42
+ "tie_word_embeddings": false,
43
+ "torch_dtype": "bfloat16",
44
+ "transformers_version": "4.31.0",
45
+ "tune_mm_mlp_adapter": false,
46
+ "use_cache": true,
47
+ "use_mm_proj": true,
48
+ "vocab_size": 32000
49
+ }
alana-chatunivi-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/non_lora_trainables.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:60fb82c3660319e6d0b239950b20c28181e97f1ade117dc0660b40e2ad94a89b
3
+ size 912
alana-chatunivi-vqa-gemini-egoclip-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/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
alana-chatunivi-vqa-gemini-egoclip-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/adapter_config.json ADDED
@@ -0,0 +1,27 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "Chat-UniVi/Chat-UniVi",
4
+ "bias": "none",
5
+ "fan_in_fan_out": false,
6
+ "inference_mode": true,
7
+ "init_lora_weights": true,
8
+ "layers_pattern": null,
9
+ "layers_to_transform": null,
10
+ "lora_alpha": 128,
11
+ "lora_dropout": 0.05,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 64,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "o_proj",
18
+ "up_proj",
19
+ "q_proj",
20
+ "v_proj",
21
+ "gate_proj",
22
+ "down_proj",
23
+ "k_proj",
24
+ "mm_projector"
25
+ ],
26
+ "task_type": "CAUSAL_LM"
27
+ }
alana-chatunivi-vqa-gemini-egoclip-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dbb1ab68fe282b6c6c5651a68ead0785b2881fbbf2639eac0cfbf5308f98b635
3
+ size 320627474
alana-chatunivi-vqa-gemini-egoclip-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/config.json ADDED
@@ -0,0 +1,49 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "Chat-UniVi/Chat-UniVi",
3
+ "aarchitectures": "LlavaLlamaForCausalLM",
4
+ "architectures": [
5
+ "ChatUniViLlamaForCausalLM"
6
+ ],
7
+ "bos_token_id": 1,
8
+ "config": {
9
+ "freeze": false,
10
+ "mm_tune": true,
11
+ "spatial_cluster_rate0": 64,
12
+ "spatial_cluster_rate1": 32,
13
+ "spatial_cluster_rate2": 16,
14
+ "temporal_cluster_rate": 0.0625,
15
+ "use_cluster": true,
16
+ "vision_tune": false
17
+ },
18
+ "eos_token_id": 2,
19
+ "freeze_mm_mlp_adapter": false,
20
+ "hidden_act": "silu",
21
+ "hidden_size": 4096,
22
+ "image_aspect_ratio": "square",
23
+ "image_grid_pinpoints": null,
24
+ "initializer_range": 0.02,
25
+ "intermediate_size": 11008,
26
+ "max_position_embeddings": 4096,
27
+ "mm_hidden_size": 1024,
28
+ "mm_use_box_start_end": false,
29
+ "mm_use_im_patch_token": false,
30
+ "mm_use_im_start_end": false,
31
+ "mm_vision_select_feature": "patch",
32
+ "mm_vision_select_layer": -2,
33
+ "mm_vision_tower": "openai/clip-vit-large-patch14",
34
+ "model_type": "ChatUniVi",
35
+ "num_attention_heads": 32,
36
+ "num_hidden_layers": 32,
37
+ "num_key_value_heads": 32,
38
+ "pad_token_id": 0,
39
+ "pretraining_tp": 1,
40
+ "rms_norm_eps": 1e-05,
41
+ "rope_scaling": null,
42
+ "tie_word_embeddings": false,
43
+ "torch_dtype": "bfloat16",
44
+ "transformers_version": "4.31.0",
45
+ "tune_mm_mlp_adapter": false,
46
+ "use_cache": true,
47
+ "use_mm_proj": true,
48
+ "vocab_size": 32000
49
+ }
alana-chatunivi-vqa-gemini-egoclip-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/non_lora_trainables.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:60fb82c3660319e6d0b239950b20c28181e97f1ade117dc0660b40e2ad94a89b
3
+ size 912
alana-chatunivi-vqa-gemini-egoclip-hm3d-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/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
alana-chatunivi-vqa-gemini-egoclip-hm3d-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/adapter_config.json ADDED
@@ -0,0 +1,27 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "Chat-UniVi/Chat-UniVi",
4
+ "bias": "none",
5
+ "fan_in_fan_out": false,
6
+ "inference_mode": true,
7
+ "init_lora_weights": true,
8
+ "layers_pattern": null,
9
+ "layers_to_transform": null,
10
+ "lora_alpha": 128,
11
+ "lora_dropout": 0.05,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 64,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "o_proj",
18
+ "gate_proj",
19
+ "up_proj",
20
+ "q_proj",
21
+ "k_proj",
22
+ "down_proj",
23
+ "v_proj",
24
+ "mm_projector"
25
+ ],
26
+ "task_type": "CAUSAL_LM"
27
+ }
alana-chatunivi-vqa-gemini-egoclip-hm3d-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:85e528814b15947522d31a8893e4f2ac186c18b0055cb92e95269de53be56a2c
3
+ size 320627474
alana-chatunivi-vqa-gemini-egoclip-hm3d-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/config.json ADDED
@@ -0,0 +1,49 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "Chat-UniVi/Chat-UniVi",
3
+ "aarchitectures": "LlavaLlamaForCausalLM",
4
+ "architectures": [
5
+ "ChatUniViLlamaForCausalLM"
6
+ ],
7
+ "bos_token_id": 1,
8
+ "config": {
9
+ "freeze": false,
10
+ "mm_tune": true,
11
+ "spatial_cluster_rate0": 64,
12
+ "spatial_cluster_rate1": 32,
13
+ "spatial_cluster_rate2": 16,
14
+ "temporal_cluster_rate": 0.0625,
15
+ "use_cluster": true,
16
+ "vision_tune": false
17
+ },
18
+ "eos_token_id": 2,
19
+ "freeze_mm_mlp_adapter": false,
20
+ "hidden_act": "silu",
21
+ "hidden_size": 4096,
22
+ "image_aspect_ratio": "square",
23
+ "image_grid_pinpoints": null,
24
+ "initializer_range": 0.02,
25
+ "intermediate_size": 11008,
26
+ "max_position_embeddings": 4096,
27
+ "mm_hidden_size": 1024,
28
+ "mm_use_box_start_end": false,
29
+ "mm_use_im_patch_token": false,
30
+ "mm_use_im_start_end": false,
31
+ "mm_vision_select_feature": "patch",
32
+ "mm_vision_select_layer": -2,
33
+ "mm_vision_tower": "openai/clip-vit-large-patch14",
34
+ "model_type": "ChatUniVi",
35
+ "num_attention_heads": 32,
36
+ "num_hidden_layers": 32,
37
+ "num_key_value_heads": 32,
38
+ "pad_token_id": 0,
39
+ "pretraining_tp": 1,
40
+ "rms_norm_eps": 1e-05,
41
+ "rope_scaling": null,
42
+ "tie_word_embeddings": false,
43
+ "torch_dtype": "bfloat16",
44
+ "transformers_version": "4.31.0",
45
+ "tune_mm_mlp_adapter": false,
46
+ "use_cache": true,
47
+ "use_mm_proj": true,
48
+ "vocab_size": 32000
49
+ }
alana-chatunivi-vqa-gemini-egoclip-hm3d-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/non_lora_trainables.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:60fb82c3660319e6d0b239950b20c28181e97f1ade117dc0660b40e2ad94a89b
3
+ size 912
alana-chatunivi-vqa-gemini-vsr-egoclip-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/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
alana-chatunivi-vqa-gemini-vsr-egoclip-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/adapter_config.json ADDED
@@ -0,0 +1,27 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "Chat-UniVi/Chat-UniVi",
4
+ "bias": "none",
5
+ "fan_in_fan_out": false,
6
+ "inference_mode": true,
7
+ "init_lora_weights": true,
8
+ "layers_pattern": null,
9
+ "layers_to_transform": null,
10
+ "lora_alpha": 128,
11
+ "lora_dropout": 0.05,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 64,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "o_proj",
18
+ "k_proj",
19
+ "q_proj",
20
+ "gate_proj",
21
+ "mm_projector",
22
+ "up_proj",
23
+ "down_proj",
24
+ "v_proj"
25
+ ],
26
+ "task_type": "CAUSAL_LM"
27
+ }
alana-chatunivi-vqa-gemini-vsr-egoclip-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:92271c5862c0183372726069df3ec7e2f022d7eb81547dfc2d1a1dada7730ccc
3
+ size 320627474
alana-chatunivi-vqa-gemini-vsr-egoclip-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/config.json ADDED
@@ -0,0 +1,49 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "Chat-UniVi/Chat-UniVi",
3
+ "aarchitectures": "LlavaLlamaForCausalLM",
4
+ "architectures": [
5
+ "ChatUniViLlamaForCausalLM"
6
+ ],
7
+ "bos_token_id": 1,
8
+ "config": {
9
+ "freeze": false,
10
+ "mm_tune": true,
11
+ "spatial_cluster_rate0": 64,
12
+ "spatial_cluster_rate1": 32,
13
+ "spatial_cluster_rate2": 16,
14
+ "temporal_cluster_rate": 0.0625,
15
+ "use_cluster": true,
16
+ "vision_tune": false
17
+ },
18
+ "eos_token_id": 2,
19
+ "freeze_mm_mlp_adapter": false,
20
+ "hidden_act": "silu",
21
+ "hidden_size": 4096,
22
+ "image_aspect_ratio": "square",
23
+ "image_grid_pinpoints": null,
24
+ "initializer_range": 0.02,
25
+ "intermediate_size": 11008,
26
+ "max_position_embeddings": 4096,
27
+ "mm_hidden_size": 1024,
28
+ "mm_use_box_start_end": false,
29
+ "mm_use_im_patch_token": false,
30
+ "mm_use_im_start_end": false,
31
+ "mm_vision_select_feature": "patch",
32
+ "mm_vision_select_layer": -2,
33
+ "mm_vision_tower": "openai/clip-vit-large-patch14",
34
+ "model_type": "ChatUniVi",
35
+ "num_attention_heads": 32,
36
+ "num_hidden_layers": 32,
37
+ "num_key_value_heads": 32,
38
+ "pad_token_id": 0,
39
+ "pretraining_tp": 1,
40
+ "rms_norm_eps": 1e-05,
41
+ "rope_scaling": null,
42
+ "tie_word_embeddings": false,
43
+ "torch_dtype": "bfloat16",
44
+ "transformers_version": "4.31.0",
45
+ "tune_mm_mlp_adapter": false,
46
+ "use_cache": true,
47
+ "use_mm_proj": true,
48
+ "vocab_size": 32000
49
+ }
alana-chatunivi-vqa-gemini-vsr-egoclip-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/non_lora_trainables.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:60fb82c3660319e6d0b239950b20c28181e97f1ade117dc0660b40e2ad94a89b
3
+ size 912
alana-chatunivi-vqa-gemini-vsr-egoclip-hm3d-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/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
alana-chatunivi-vqa-gemini-vsr-egoclip-hm3d-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/adapter_config.json ADDED
@@ -0,0 +1,27 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "Chat-UniVi/Chat-UniVi",
4
+ "bias": "none",
5
+ "fan_in_fan_out": false,
6
+ "inference_mode": true,
7
+ "init_lora_weights": true,
8
+ "layers_pattern": null,
9
+ "layers_to_transform": null,
10
+ "lora_alpha": 128,
11
+ "lora_dropout": 0.05,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 64,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "gate_proj",
18
+ "mm_projector",
19
+ "o_proj",
20
+ "k_proj",
21
+ "up_proj",
22
+ "down_proj",
23
+ "v_proj",
24
+ "q_proj"
25
+ ],
26
+ "task_type": "CAUSAL_LM"
27
+ }
alana-chatunivi-vqa-gemini-vsr-egoclip-hm3d-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7c271a1fcca32f7c6609cd938d63c6f2fca0d64492ccc4370da2025fbf38f88c
3
+ size 320627474
alana-chatunivi-vqa-gemini-vsr-egoclip-hm3d-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/config.json ADDED
@@ -0,0 +1,49 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "Chat-UniVi/Chat-UniVi",
3
+ "aarchitectures": "LlavaLlamaForCausalLM",
4
+ "architectures": [
5
+ "ChatUniViLlamaForCausalLM"
6
+ ],
7
+ "bos_token_id": 1,
8
+ "config": {
9
+ "freeze": false,
10
+ "mm_tune": true,
11
+ "spatial_cluster_rate0": 64,
12
+ "spatial_cluster_rate1": 32,
13
+ "spatial_cluster_rate2": 16,
14
+ "temporal_cluster_rate": 0.0625,
15
+ "use_cluster": true,
16
+ "vision_tune": false
17
+ },
18
+ "eos_token_id": 2,
19
+ "freeze_mm_mlp_adapter": false,
20
+ "hidden_act": "silu",
21
+ "hidden_size": 4096,
22
+ "image_aspect_ratio": "square",
23
+ "image_grid_pinpoints": null,
24
+ "initializer_range": 0.02,
25
+ "intermediate_size": 11008,
26
+ "max_position_embeddings": 4096,
27
+ "mm_hidden_size": 1024,
28
+ "mm_use_box_start_end": false,
29
+ "mm_use_im_patch_token": false,
30
+ "mm_use_im_start_end": false,
31
+ "mm_vision_select_feature": "patch",
32
+ "mm_vision_select_layer": -2,
33
+ "mm_vision_tower": "openai/clip-vit-large-patch14",
34
+ "model_type": "ChatUniVi",
35
+ "num_attention_heads": 32,
36
+ "num_hidden_layers": 32,
37
+ "num_key_value_heads": 32,
38
+ "pad_token_id": 0,
39
+ "pretraining_tp": 1,
40
+ "rms_norm_eps": 1e-05,
41
+ "rope_scaling": null,
42
+ "tie_word_embeddings": false,
43
+ "torch_dtype": "bfloat16",
44
+ "transformers_version": "4.31.0",
45
+ "tune_mm_mlp_adapter": false,
46
+ "use_cache": true,
47
+ "use_mm_proj": true,
48
+ "vocab_size": 32000
49
+ }
alana-chatunivi-vqa-gemini-vsr-egoclip-hm3d-acc64-lr3e4-warmup_ratio025-weight_decay001-lora_r64-lora_alpha128/non_lora_trainables.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:60fb82c3660319e6d0b239950b20c28181e97f1ade117dc0660b40e2ad94a89b
3
+ size 912