afonsosamarques
commited on
Commit
•
f5baa31
1
Parent(s):
627769f
Training in progress, step 2000
Browse files- .gitignore +1 -0
- config.json +83 -0
- pytorch_model.bin +3 -0
- training_args.bin +3 -0
.gitignore
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
checkpoint-*/
|
config.json
ADDED
@@ -0,0 +1,83 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"act_dim": 6,
|
3 |
+
"action_tanh": true,
|
4 |
+
"activation_function": "relu",
|
5 |
+
"adv_act_dim": 6,
|
6 |
+
"architectures": [
|
7 |
+
"VanillaDT"
|
8 |
+
],
|
9 |
+
"attn_pdrop": 0.1,
|
10 |
+
"bos_token_id": 50256,
|
11 |
+
"context_size": 20,
|
12 |
+
"embd_pdrop": 0.1,
|
13 |
+
"eos_token_id": 50256,
|
14 |
+
"flag": false,
|
15 |
+
"hidden_size": 128,
|
16 |
+
"initializer_range": 0.02,
|
17 |
+
"lambda1": 1.0,
|
18 |
+
"lambda2": 1.0,
|
19 |
+
"layer_norm_epsilon": 1e-05,
|
20 |
+
"log_interval_steps": 100,
|
21 |
+
"max_ep_len": 999,
|
22 |
+
"max_ep_return": 1289,
|
23 |
+
"max_obs_len": 999,
|
24 |
+
"max_obs_return": 1289,
|
25 |
+
"min_ep_return": -96.82057464200004,
|
26 |
+
"min_obs_return": -96.82057464200004,
|
27 |
+
"model_type": "decision_transformer",
|
28 |
+
"n_head": 1,
|
29 |
+
"n_inner": null,
|
30 |
+
"n_layer": 3,
|
31 |
+
"n_positions": 1024,
|
32 |
+
"pr_act_dim": 6,
|
33 |
+
"reorder_and_upcast_attn": false,
|
34 |
+
"resid_pdrop": 0.1,
|
35 |
+
"returns_scale": 1000,
|
36 |
+
"scale_attn_by_inverse_layer_idx": false,
|
37 |
+
"scale_attn_weights": true,
|
38 |
+
"state_dim": 17,
|
39 |
+
"state_mean": [
|
40 |
+
1.1575770125674947,
|
41 |
+
0.16765135255888688,
|
42 |
+
-0.10125198559126013,
|
43 |
+
-0.048161959123527304,
|
44 |
+
0.5866500870318963,
|
45 |
+
-0.08899718695545204,
|
46 |
+
-0.061814936780544304,
|
47 |
+
0.4317695335151212,
|
48 |
+
0.6883041465189006,
|
49 |
+
-0.16324784546275664,
|
50 |
+
0.19899619193857912,
|
51 |
+
-0.20914407994362408,
|
52 |
+
-0.13819401972845488,
|
53 |
+
-0.16121668431453506,
|
54 |
+
-0.18814624439028538,
|
55 |
+
-0.22432487455365363,
|
56 |
+
-0.4176328378528237
|
57 |
+
],
|
58 |
+
"state_std": [
|
59 |
+
0.07913902740370016,
|
60 |
+
0.31899006345311365,
|
61 |
+
0.2802079737062644,
|
62 |
+
0.21292672390975065,
|
63 |
+
0.4762548011641742,
|
64 |
+
0.255352584176118,
|
65 |
+
0.28210621331296115,
|
66 |
+
0.6152170140751688,
|
67 |
+
0.845679173234231,
|
68 |
+
0.45206512647207403,
|
69 |
+
1.7121676393489746,
|
70 |
+
1.8135847754595518,
|
71 |
+
2.064758437346931,
|
72 |
+
2.6553273730109574,
|
73 |
+
1.7935814824240828,
|
74 |
+
1.9478784717081086,
|
75 |
+
2.8018697185903454
|
76 |
+
],
|
77 |
+
"torch_dtype": "float32",
|
78 |
+
"total_train_steps": 10000,
|
79 |
+
"transformers_version": "4.29.2",
|
80 |
+
"use_cache": true,
|
81 |
+
"vocab_size": 1,
|
82 |
+
"warmup_steps": 1000
|
83 |
+
}
|
pytorch_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:2c120d743db822807ff31c291076c1c4ab4cebbce3baea4568e683e216d270f6
|
3 |
+
size 6608997
|
training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:efef5d0a3308fd97c3bdc3b1328db8eefa09688659ae1aed294bdff8d2c76920
|
3 |
+
size 4664
|