Spaces:
Running
on
Zero
Running
on
Zero
add checkpoint
Browse files
.DS_Store
ADDED
Binary file (6.15 kB). View file
|
|
data/.DS_Store
ADDED
Binary file (6.15 kB). View file
|
|
data/mar_ckpt/.DS_Store
ADDED
Binary file (6.15 kB). View file
|
|
data/mar_ckpt/langtable/.DS_Store
ADDED
Binary file (6.15 kB). View file
|
|
data/mar_ckpt/langtable/README.md
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
tags:
|
3 |
+
- model_hub_mixin
|
4 |
+
- pytorch_model_hub_mixin
|
5 |
+
---
|
6 |
+
|
7 |
+
This model has been pushed to the Hub using the [PytorchModelHubMixin](https://huggingface.co/docs/huggingface_hub/package_reference/mixins#huggingface_hub.PyTorchModelHubMixin) integration:
|
8 |
+
- Library: [More Information Needed]
|
9 |
+
- Docs: [More Information Needed]
|
data/mar_ckpt/langtable/config.json
ADDED
@@ -0,0 +1,72 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"Diffusion": true,
|
3 |
+
"S": 1024,
|
4 |
+
"T": 12,
|
5 |
+
"action_contrastive_loss": false,
|
6 |
+
"action_domains": [
|
7 |
+
"language_table"
|
8 |
+
],
|
9 |
+
"action_loss_weight": 0.5,
|
10 |
+
"action_network": "conconcat+modulatecat",
|
11 |
+
"action_stats": [
|
12 |
+
[
|
13 |
+
[
|
14 |
+
0.00014842326345387846,
|
15 |
+
-0.0005635050474666059
|
16 |
+
],
|
17 |
+
[
|
18 |
+
0.030163198709487915,
|
19 |
+
0.042305462062358856
|
20 |
+
]
|
21 |
+
]
|
22 |
+
],
|
23 |
+
"action_token_size": 64,
|
24 |
+
"arch": "STTransformerDecoder",
|
25 |
+
"attn_drop": 0.1,
|
26 |
+
"attn_dropout": 0.1,
|
27 |
+
"buffer_size": 64,
|
28 |
+
"d_action": 28,
|
29 |
+
"d_actions": [
|
30 |
+
2
|
31 |
+
],
|
32 |
+
"d_model": 256,
|
33 |
+
"dataloader_apply_corruption": false,
|
34 |
+
"dataloader_apply_mask": true,
|
35 |
+
"dataloader_mask_ratio_min": 0.1,
|
36 |
+
"diffloss_d": 4,
|
37 |
+
"diffloss_w": 1024,
|
38 |
+
"diffusion_batch_mul": 1,
|
39 |
+
"dim": 512,
|
40 |
+
"drop_action_ratio": 0.0,
|
41 |
+
"factored_vocab_size": 512,
|
42 |
+
"grad_checkpointing": false,
|
43 |
+
"image_vocab_size": null,
|
44 |
+
"init_actions": true,
|
45 |
+
"jointly_predict_actions": false,
|
46 |
+
"jointly_predict_states": true,
|
47 |
+
"label_drop_prob": 0.5,
|
48 |
+
"mask_ratio_min": 0.7,
|
49 |
+
"maskgit_steps": 16,
|
50 |
+
"max_corrupt_rate": 0.2,
|
51 |
+
"mlp_bias": false,
|
52 |
+
"mlp_drop": 0.05,
|
53 |
+
"mlp_ratio": 4.0,
|
54 |
+
"non_mlm_ratio": 0.2,
|
55 |
+
"num_factored_vocabs": 2,
|
56 |
+
"num_heads": 8,
|
57 |
+
"num_layers": 32,
|
58 |
+
"num_prompt_frames": 4,
|
59 |
+
"num_sampling_steps": "100",
|
60 |
+
"patch_size": 2,
|
61 |
+
"predict_unmask": false,
|
62 |
+
"proj_bias": true,
|
63 |
+
"proj_dropout": 0.1,
|
64 |
+
"qk_norm": false,
|
65 |
+
"qkv_bias": true,
|
66 |
+
"random_dummy_action": true,
|
67 |
+
"shared_action_mlps": true,
|
68 |
+
"use_actions": true,
|
69 |
+
"use_mup": false,
|
70 |
+
"vae_embed_dim": 4,
|
71 |
+
"vae_stride": 1
|
72 |
+
}
|
data/mar_ckpt/langtable/random_states_0.pkl
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:67d746c0889464b9285190f8f0149e64abd026b74e771f5e0ba9ca43d50e585a
|
3 |
+
size 16100
|
data/mar_ckpt/langtable/scheduler.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9d394254995d1bcdcbb47584b236a9c1cde474fb49c5670dacddbb7ece6a727f
|
3 |
+
size 1064
|