Commit
•
fda8776
1
Parent(s):
15a18ba
Training in progress, step 2000
Browse files- .gitignore +1 -0
- config.json +71 -0
- pytorch_model.bin +3 -0
- training_args.bin +3 -0
.gitignore
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
checkpoint-*/
|
config.json
ADDED
|
@@ -0,0 +1,71 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"act_dim": 3,
|
| 3 |
+
"action_tanh": true,
|
| 4 |
+
"activation_function": "relu",
|
| 5 |
+
"adv_act_dim": 3,
|
| 6 |
+
"architectures": [
|
| 7 |
+
"AdversarialDT"
|
| 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": 3073,
|
| 23 |
+
"max_obs_len": 999,
|
| 24 |
+
"max_obs_return": 3073,
|
| 25 |
+
"min_ep_return": 0.98062783,
|
| 26 |
+
"min_obs_return": 0.98062783,
|
| 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": 3,
|
| 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": 11,
|
| 39 |
+
"state_mean": [
|
| 40 |
+
1.209087606182348,
|
| 41 |
+
-0.0024433804464966537,
|
| 42 |
+
-0.5121224035046005,
|
| 43 |
+
-0.1792276788145927,
|
| 44 |
+
0.10206460331576518,
|
| 45 |
+
1.9267843346155455,
|
| 46 |
+
-0.18265996445338303,
|
| 47 |
+
0.02094631569481702,
|
| 48 |
+
-0.3570970607279206,
|
| 49 |
+
-0.25557554713170616,
|
| 50 |
+
-0.5014115326841497
|
| 51 |
+
],
|
| 52 |
+
"state_std": [
|
| 53 |
+
0.14455136461248416,
|
| 54 |
+
0.08059303935274553,
|
| 55 |
+
0.2651650281434884,
|
| 56 |
+
0.2943715715574719,
|
| 57 |
+
0.6045520274648529,
|
| 58 |
+
0.9001491436425907,
|
| 59 |
+
1.2017406710493554,
|
| 60 |
+
1.029395608366261,
|
| 61 |
+
2.059115337608456,
|
| 62 |
+
2.898265052736193,
|
| 63 |
+
6.197472424851261
|
| 64 |
+
],
|
| 65 |
+
"torch_dtype": "float32",
|
| 66 |
+
"total_train_steps": 10000,
|
| 67 |
+
"transformers_version": "4.29.2",
|
| 68 |
+
"use_cache": true,
|
| 69 |
+
"vocab_size": 1,
|
| 70 |
+
"warmup_steps": 1000
|
| 71 |
+
}
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0f2977e42233c532961f8dd760da633a3c371fd70bc51afa0fa80be602eded3e
|
| 3 |
+
size 6609405
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ce49bf17c1b704aa32bf1fcb9168c547f8604e1c36868bcdc58452f9677459e6
|
| 3 |
+
size 4664
|