Hugging Face's logo

Stanford-ILIAD
/
minivla-vq-libero90-prismatic

Image-Text-to-Text
Transformers
English
robotics
vla
multimodal
pretraining
Inference Endpoints
Model card Files Community
1
belkhale commited on
Commit
cb0f910
•
1 Parent(s): de08793

Upload folder using huggingface_hub

Browse files
Files changed (8) hide show
  1. .gitattributes +1 -0
  2. checkpoints/step-025000-epoch-11-loss=0.5844.pt +3 -0
  3. checkpoints/step-150000-epoch-67-loss=0.0934.pt +3 -0
  4. config.json +46 -0
  5. config.yaml +42 -0
  6. dataset_statistics.json +127 -0
  7. prism-qwen25-dinosiglip-224px+0_5b+mx-libero-90+n1+b32+x7--vq_extra_tokenizer.jsonl +3 -0
  8. run-metrics.jsonl +1 -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
+ prism-qwen25-dinosiglip-224px+0_5b+mx-libero-90+n1+b32+x7--vq_extra_tokenizer.jsonl filter=lfs diff=lfs merge=lfs -text
checkpoints/step-025000-epoch-11-loss=0.5844.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9b08987019dc1f2ab5dfa0d81fd5343c37062e37704c26fb637e63c0fa07fe18
3
+ size 5554882540
checkpoints/step-150000-epoch-67-loss=0.0934.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e99cfd157acca023e10b6b2d0305a2e4d480ff55a371646012404e9b32c3700b
3
+ size 5554882540
config.json ADDED
@@ -0,0 +1,46 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "data_root_dir": "/hai/scratch/belkhale/datasets",
3
+ "hf_token": ".hf_token",
4
+ "image_aug": false,
5
+ "is_resume": true,
6
+ "pretrained_checkpoint": "runs/prism-qwen25-dinosiglip-224px+0_5b+mx-libero-90+n1+b32+x7--vq_extra_tokenizer/checkpoints/step-065000-epoch-29-loss=0.4407.pt",
7
+ "resume_epoch": 29,
8
+ "resume_step": 65000,
9
+ "run_id": "prism-qwen25-dinosiglip-224px+0_5b+mx-libero-90+n1+b32+x7--vq_extra_tokenizer",
10
+ "run_id_note": "vq_extra_tokenizer",
11
+ "run_root_dir": "runs",
12
+ "save_interval": 2500,
13
+ "seed": 7,
14
+ "trackers": [
15
+ "jsonl",
16
+ "wandb"
17
+ ],
18
+ "vla": {
19
+ "action_tokenizer": "libero_vq_extra_action_tokenizer",
20
+ "base_vlm": "prism-qwen25-extra-dinosiglip-224px+0_5b",
21
+ "data_mix": "libero_90",
22
+ "enable_gradient_checkpointing": true,
23
+ "enable_mixed_precision_training": true,
24
+ "epochs": 1000,
25
+ "expected_world_size": 8,
26
+ "freeze_llm_backbone": false,
27
+ "freeze_vision_backbone": false,
28
+ "global_batch_size": 256,
29
+ "learning_rate": 2e-05,
30
+ "lr_scheduler_type": "constant",
31
+ "max_grad_norm": 1.0,
32
+ "max_steps": null,
33
+ "per_device_batch_size": 32,
34
+ "reduce_in_full_precision": true,
35
+ "save_every_n_steps": 25000,
36
+ "shuffle_buffer_size": 256000,
37
+ "train_strategy": "fsdp-full-shard",
38
+ "type": "prism-qwen25-dinosiglip-224px+0_5b+mx-libero-90",
39
+ "unfreeze_last_llm_layer": false,
40
+ "vla_id": "prism-qwen25-dinosiglip-224px+0_5b+mx-libero-90",
41
+ "warmup_ratio": 0.0,
42
+ "weight_decay": 0.0
43
+ },
44
+ "wandb_entity": null,
45
+ "wandb_project": "prismatic"
46
+ }
config.yaml ADDED
@@ -0,0 +1,42 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ data_root_dir: /hai/scratch/belkhale/datasets
2
+ hf_token: .hf_token
3
+ image_aug: false
4
+ is_resume: true
5
+ pretrained_checkpoint: runs/prism-qwen25-dinosiglip-224px+0_5b+mx-libero-90+n1+b32+x7--vq_extra_tokenizer/checkpoints/step-065000-epoch-29-loss=0.4407.pt
6
+ resume_epoch: 29
7
+ resume_step: 65000
8
+ run_id: prism-qwen25-dinosiglip-224px+0_5b+mx-libero-90+n1+b32+x7--vq_extra_tokenizer
9
+ run_id_note: vq_extra_tokenizer
10
+ run_root_dir: runs
11
+ save_interval: 2500
12
+ seed: 7
13
+ trackers:
14
+ - jsonl
15
+ - wandb
16
+ vla:
17
+ action_tokenizer: libero_vq_extra_action_tokenizer
18
+ base_vlm: prism-qwen25-extra-dinosiglip-224px+0_5b
19
+ data_mix: libero_90
20
+ enable_gradient_checkpointing: true
21
+ enable_mixed_precision_training: true
22
+ epochs: 1000
23
+ expected_world_size: 8
24
+ freeze_llm_backbone: false
25
+ freeze_vision_backbone: false
26
+ global_batch_size: 256
27
+ learning_rate: 2.0e-05
28
+ lr_scheduler_type: constant
29
+ max_grad_norm: 1.0
30
+ max_steps: null
31
+ per_device_batch_size: 32
32
+ reduce_in_full_precision: true
33
+ save_every_n_steps: 25000
34
+ shuffle_buffer_size: 256000
35
+ train_strategy: fsdp-full-shard
36
+ type: prism-qwen25-dinosiglip-224px+0_5b+mx-libero-90
37
+ unfreeze_last_llm_layer: false
38
+ vla_id: prism-qwen25-dinosiglip-224px+0_5b+mx-libero-90
39
+ warmup_ratio: 0.0
40
+ weight_decay: 0.0
41
+ wandb_entity: null
42
+ wandb_project: prismatic
dataset_statistics.json ADDED
@@ -0,0 +1,127 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "libero_90": {
3
+ "action": {
4
+ "mean": [
5
+ 0.04548764228820801,
6
+ 0.037047673016786575,
7
+ -0.09770790487527847,
8
+ 0.005058619659394026,
9
+ 0.00224184594117105,
10
+ -0.006234025117009878,
11
+ 0.528744101524353
12
+ ],
13
+ "std": [
14
+ 0.2989887297153473,
15
+ 0.361598402261734,
16
+ 0.4069668650627136,
17
+ 0.04840222746133804,
18
+ 0.058257583528757095,
19
+ 0.08737559616565704,
20
+ 0.49890485405921936
21
+ ],
22
+ "max": [
23
+ 0.9375,
24
+ 0.9375,
25
+ 0.9375,
26
+ 0.375,
27
+ 0.375,
28
+ 0.375,
29
+ 1.0
30
+ ],
31
+ "min": [
32
+ -0.9375,
33
+ -0.9375,
34
+ -0.9375,
35
+ -0.3257142901420593,
36
+ -0.375,
37
+ -0.375,
38
+ 0.0
39
+ ],
40
+ "q01": [
41
+ -0.6321428418159485,
42
+ -0.8732143044471741,
43
+ -0.8973214030265808,
44
+ -0.12321428209543228,
45
+ -0.15642857551574707,
46
+ -0.2807142734527588,
47
+ 0.0
48
+ ],
49
+ "q99": [
50
+ 0.8517857193946838,
51
+ 0.8464285731315613,
52
+ 0.9375,
53
+ 0.1875,
54
+ 0.1778571456670761,
55
+ 0.34928572177886963,
56
+ 1.0
57
+ ],
58
+ "mask": [
59
+ true,
60
+ true,
61
+ true,
62
+ true,
63
+ true,
64
+ true,
65
+ false
66
+ ]
67
+ },
68
+ "proprio": {
69
+ "mean": [
70
+ 0.0,
71
+ 0.0,
72
+ 0.0,
73
+ 0.0,
74
+ 0.0,
75
+ 0.0,
76
+ 0.0
77
+ ],
78
+ "std": [
79
+ 0.0,
80
+ 0.0,
81
+ 0.0,
82
+ 0.0,
83
+ 0.0,
84
+ 0.0,
85
+ 0.0
86
+ ],
87
+ "max": [
88
+ 0.0,
89
+ 0.0,
90
+ 0.0,
91
+ 0.0,
92
+ 0.0,
93
+ 0.0,
94
+ 0.0
95
+ ],
96
+ "min": [
97
+ 0.0,
98
+ 0.0,
99
+ 0.0,
100
+ 0.0,
101
+ 0.0,
102
+ 0.0,
103
+ 0.0
104
+ ],
105
+ "q01": [
106
+ 0.0,
107
+ 0.0,
108
+ 0.0,
109
+ 0.0,
110
+ 0.0,
111
+ 0.0,
112
+ 0.0
113
+ ],
114
+ "q99": [
115
+ 0.0,
116
+ 0.0,
117
+ 0.0,
118
+ 0.0,
119
+ 0.0,
120
+ 0.0,
121
+ 0.0
122
+ ]
123
+ },
124
+ "num_transitions": 569230,
125
+ "num_trajectories": 3924
126
+ }
127
+ }
prism-qwen25-dinosiglip-224px+0_5b+mx-libero-90+n1+b32+x7--vq_extra_tokenizer.jsonl ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:80952c0854fd4216968dbbbd72b52187a28a79fc4996f05b55fa99cb234edb35
3
+ size 50204058
run-metrics.jsonl ADDED
@@ -0,0 +1 @@
 
 
1
+ {"hparams": {"data_root_dir": "/hai/scratch/belkhale/datasets", "hf_token": ".hf_token", "image_aug": false, "is_resume": true, "pretrained_checkpoint": "runs/prism-qwen25-dinosiglip-224px+0_5b+mx-libero-90+n1+b32+x7--vq_extra_tokenizer/checkpoints/step-065000-epoch-29-loss=0.4407.pt", "resume_epoch": 29, "resume_step": 65000, "run_id": "prism-qwen25-dinosiglip-224px+0_5b+mx-libero-90+n1+b32+x7--vq_extra_tokenizer", "run_id_note": "vq_extra_tokenizer", "run_root_dir": "runs", "save_interval": 2500, "seed": 7, "trackers": ["jsonl", "wandb"], "vla": {"action_tokenizer": "libero_vq_extra_action_tokenizer", "base_vlm": "prism-qwen25-extra-dinosiglip-224px+0_5b", "data_mix": "libero_90", "enable_gradient_checkpointing": true, "enable_mixed_precision_training": true, "epochs": 1000, "expected_world_size": 8, "freeze_llm_backbone": false, "freeze_vision_backbone": false, "global_batch_size": 256, "learning_rate": 2e-05, "lr_scheduler_type": "constant", "max_grad_norm": 1.0, "max_steps": null, "per_device_batch_size": 32, "reduce_in_full_precision": true, "save_every_n_steps": 25000, "shuffle_buffer_size": 256000, "train_strategy": "fsdp-full-shard", "type": "prism-qwen25-dinosiglip-224px+0_5b+mx-libero-90", "unfreeze_last_llm_layer": false, "vla_id": "prism-qwen25-dinosiglip-224px+0_5b+mx-libero-90", "warmup_ratio": 0.0, "weight_decay": 0.0}, "wandb_entity": null, "wandb_project": "prismatic"}, "run_id": "prism-qwen25-dinosiglip-224px+0_5b+mx-libero-90+n1+b32+x7--vq_extra_tokenizer"}