Upload folder using huggingface_hub
Browse files- large-rl/ckpts/policy_weights/policy_epoch_0.pt +3 -0
- large-rl/ckpts/policy_weights/policy_epoch_10.pt +3 -0
- large-rl/ckpts/policy_weights/policy_epoch_12.pt +3 -0
- large-rl/ckpts/policy_weights/policy_epoch_14.pt +3 -0
- large-rl/ckpts/policy_weights/policy_epoch_16.pt +3 -0
- large-rl/ckpts/policy_weights/policy_epoch_18.pt +3 -0
- large-rl/ckpts/policy_weights/policy_epoch_2.pt +3 -0
- large-rl/ckpts/policy_weights/policy_epoch_20.pt +3 -0
- large-rl/ckpts/policy_weights/policy_epoch_22.pt +3 -0
- large-rl/ckpts/policy_weights/policy_epoch_24.pt +3 -0
- large-rl/ckpts/policy_weights/policy_epoch_26.pt +3 -0
- large-rl/ckpts/policy_weights/policy_epoch_28.pt +3 -0
- large-rl/ckpts/policy_weights/policy_epoch_30.pt +3 -0
- large-rl/ckpts/policy_weights/policy_epoch_32.pt +3 -0
- large-rl/ckpts/policy_weights/policy_epoch_34.pt +3 -0
- large-rl/ckpts/policy_weights/policy_epoch_36.pt +3 -0
- large-rl/ckpts/policy_weights/policy_epoch_38.pt +3 -0
- large-rl/ckpts/policy_weights/policy_epoch_4.pt +3 -0
- large-rl/ckpts/policy_weights/policy_epoch_40.pt +3 -0
- large-rl/ckpts/policy_weights/policy_epoch_42.pt +3 -0
- large-rl/ckpts/policy_weights/policy_epoch_44.pt +3 -0
- large-rl/ckpts/policy_weights/policy_epoch_46.pt +3 -0
- large-rl/ckpts/policy_weights/policy_epoch_48.pt +3 -0
- large-rl/ckpts/policy_weights/policy_epoch_50.pt +3 -0
- large-rl/ckpts/policy_weights/policy_epoch_6.pt +3 -0
- large-rl/ckpts/policy_weights/policy_epoch_8.pt +3 -0
- large-rl/config.txt +106 -0
large-rl/ckpts/policy_weights/policy_epoch_0.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ee55fd734b1c5b2db4db172579704cd234528b064c3c3e0cbfcaf32fdf2f1de3
|
3 |
+
size 779315798
|
large-rl/ckpts/policy_weights/policy_epoch_10.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f4d5e0199cc56e4cf94ea5a911886c82280812a5cbbf2a98cfcb94d98fccd37d
|
3 |
+
size 779316162
|
large-rl/ckpts/policy_weights/policy_epoch_12.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:3c34e7dbf240996993d76d980dd57b89a4288db40920997c5dfb06e55bef3195
|
3 |
+
size 779316162
|
large-rl/ckpts/policy_weights/policy_epoch_14.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a3d390c0ae94e0341556348255f05b170c3d6a40b6f0e6f1d65d3587b67fd241
|
3 |
+
size 779316162
|
large-rl/ckpts/policy_weights/policy_epoch_16.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e7b63134f54bf2130105ffebfe8fe9a688bdb267ae6f32afcb8d1fbd9a15ab24
|
3 |
+
size 779316162
|
large-rl/ckpts/policy_weights/policy_epoch_18.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f50994f7f2265f06ec9bf601dc843a56f1a3d0c50a4cf48f4a2c476e02accfa7
|
3 |
+
size 779316162
|
large-rl/ckpts/policy_weights/policy_epoch_2.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ebed51f273770d3e5325e5e63bf5d72659c322c0c8559ea77f726661a52ba5f9
|
3 |
+
size 779315798
|
large-rl/ckpts/policy_weights/policy_epoch_20.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:1f6e723b842dab15131f8a09764c5b11f4b2fec4c7f54c6ab04cc56b7aa95e81
|
3 |
+
size 779316162
|
large-rl/ckpts/policy_weights/policy_epoch_22.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9a904052fef27aaad3b7db06fadb9b3784b6f9e622be2aaf9136270569f7eb88
|
3 |
+
size 779316162
|
large-rl/ckpts/policy_weights/policy_epoch_24.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:17d026917da6e9019ff730ef41e9ec5a9957f4f01b3b7153f97ad30e08e7d43e
|
3 |
+
size 779316162
|
large-rl/ckpts/policy_weights/policy_epoch_26.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f31e87a62fca3746e0e8c9d3282b66741de9a16a47cf6869ddb088893cc6bdf7
|
3 |
+
size 779316162
|
large-rl/ckpts/policy_weights/policy_epoch_28.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:fa58eb572f8ef8bb8f9551e1a1e412a1c962ad8d47a5e16b8bcc4540d7846205
|
3 |
+
size 779316162
|
large-rl/ckpts/policy_weights/policy_epoch_30.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:51556fe6cdd3a2584cc99d09d95a3d44b450024b33c6da7d40b7fd4faada1a2d
|
3 |
+
size 779316162
|
large-rl/ckpts/policy_weights/policy_epoch_32.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:59b2317b0b529c0e0a80d72f88e6b949cfb223bbc5fbeefc872d47acef105758
|
3 |
+
size 779316162
|
large-rl/ckpts/policy_weights/policy_epoch_34.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:29a0acaff33efc56d11d16311d4630496f584bc2024014e45e4ffc8b3618a8ad
|
3 |
+
size 779316162
|
large-rl/ckpts/policy_weights/policy_epoch_36.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:515573885e624ea11c3bd7e774084ac3c30da11cac2aa5a097d49e7bcc070fbf
|
3 |
+
size 779316162
|
large-rl/ckpts/policy_weights/policy_epoch_38.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:fa6534be0f29db6be5c36198f11265556e6f156c47b120d3c4fa9777167b98cc
|
3 |
+
size 779316162
|
large-rl/ckpts/policy_weights/policy_epoch_4.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f363837a845185db2ed463c9683809bf9072d593de678c6d942657d0e555820f
|
3 |
+
size 779315798
|
large-rl/ckpts/policy_weights/policy_epoch_40.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:04ed6d5179a0fbc1254384a8199cbf09eed56982c41adf0e22465bfe95044d59
|
3 |
+
size 779316162
|
large-rl/ckpts/policy_weights/policy_epoch_42.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:1ad4d8d069a8d8d11ed9792ab386efe8429715f54d79b5ec8a76084c0a120cb9
|
3 |
+
size 779316162
|
large-rl/ckpts/policy_weights/policy_epoch_44.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ba7ed5f36ab0ba07e300b190c4733adc9ad72a6b648aa5d0d1368469f45394e5
|
3 |
+
size 779316162
|
large-rl/ckpts/policy_weights/policy_epoch_46.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:1c53385d69d6c72906db383720ecaf750b83a0b6197def12dfe73f7fd35bd8e7
|
3 |
+
size 779316162
|
large-rl/ckpts/policy_weights/policy_epoch_48.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5e655f97a32bbb6afc6958f2cd9e6f8c20fc3c2b910abf06e2a667f4acf75eff
|
3 |
+
size 779316162
|
large-rl/ckpts/policy_weights/policy_epoch_50.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:1cd649d2f252a77a80f72ad5d6adbdf8517b80224f301e1fb59f06829b515a09
|
3 |
+
size 779316162
|
large-rl/ckpts/policy_weights/policy_epoch_6.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5ac4df2344b456681d89b441df91d81f5098729db545c96efdf790acf94b754c
|
3 |
+
size 779315798
|
large-rl/ckpts/policy_weights/policy_epoch_8.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e60a65f3b7ae50e1ce0c564b78d053931569a5905438f8ca9b861bb2820eebe8
|
3 |
+
size 779315798
|
large-rl/config.txt
ADDED
@@ -0,0 +1,106 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# Parameters for Actor:
|
2 |
+
# ==============================================================================
|
3 |
+
Actor.activation = 'leaky_relu'
|
4 |
+
Actor.cont_dist_kind = 'normal'
|
5 |
+
Actor.d_hidden = 512
|
6 |
+
Actor.dropout_p = 0.0
|
7 |
+
Actor.gmm_modes = 5
|
8 |
+
Actor.log_std_high = 2.0
|
9 |
+
Actor.log_std_low = -5.0
|
10 |
+
Actor.n_layers = 2
|
11 |
+
|
12 |
+
# Parameters for Agent:
|
13 |
+
# ==============================================================================
|
14 |
+
Agent.fake_filter = False
|
15 |
+
Agent.gamma = 0.999
|
16 |
+
Agent.num_critics = 4
|
17 |
+
Agent.num_critics_td = 2
|
18 |
+
Agent.offline_coeff = 1.0
|
19 |
+
Agent.online_coeff = 0.0
|
20 |
+
Agent.popart = True
|
21 |
+
Agent.reward_multiplier = 10.0
|
22 |
+
Agent.tau = 0.003
|
23 |
+
Agent.use_multigamma = True
|
24 |
+
Agent.use_target_actor = True
|
25 |
+
|
26 |
+
# Parameters for Experiment:
|
27 |
+
# ==============================================================================
|
28 |
+
Experiment.batches_per_update = 1
|
29 |
+
Experiment.critic_loss_weight = 10.0
|
30 |
+
Experiment.env_mode = 'async'
|
31 |
+
Experiment.force_reset_train_envs_every = None
|
32 |
+
Experiment.grad_clip = 1.0
|
33 |
+
Experiment.has_replay_buffer_rights = True
|
34 |
+
Experiment.l2_coeff = 0.001
|
35 |
+
Experiment.learning_rate = 0.0001
|
36 |
+
Experiment.local_time_optimizer = False
|
37 |
+
Experiment.lr_warmup_steps = 500
|
38 |
+
Experiment.mixed_precision = 'no'
|
39 |
+
Experiment.padded_sampling = 'none'
|
40 |
+
Experiment.save_trajs_as = 'npz'
|
41 |
+
Experiment.stagger_traj_file_lengths = True
|
42 |
+
Experiment.wandb_group_name = None
|
43 |
+
|
44 |
+
# Parameters for FlashAttention:
|
45 |
+
# ==============================================================================
|
46 |
+
FlashAttention.window_size = (-1, -1)
|
47 |
+
|
48 |
+
# Parameters for MetamonTstepEncoder:
|
49 |
+
# ==============================================================================
|
50 |
+
MetamonTstepEncoder.d_model = 160
|
51 |
+
MetamonTstepEncoder.extra_emb_dim = 18
|
52 |
+
MetamonTstepEncoder.n_heads = 8
|
53 |
+
MetamonTstepEncoder.n_layers = 5
|
54 |
+
MetamonTstepEncoder.scratch_tokens = 11
|
55 |
+
MetamonTstepEncoder.token_mask_aug = False
|
56 |
+
|
57 |
+
# Parameters for Multigammas:
|
58 |
+
# ==============================================================================
|
59 |
+
Multigammas.continuous = [0.1, 0.9, 0.95, 0.97, 0.99, 0.995]
|
60 |
+
Multigammas.discrete = [0.1, 0.9, 0.95, 0.97, 0.99, 0.995]
|
61 |
+
|
62 |
+
# Parameters for MultiModalEmbedding:
|
63 |
+
# ==============================================================================
|
64 |
+
MultiModalEmbedding.dropout = 0.05
|
65 |
+
MultiModalEmbedding.numerical_tokens = 6
|
66 |
+
|
67 |
+
# Parameters for NCritics:
|
68 |
+
# ==============================================================================
|
69 |
+
NCritics.activation = 'leaky_relu'
|
70 |
+
NCritics.d_hidden = 512
|
71 |
+
NCritics.dropout_p = 0.0
|
72 |
+
NCritics.n_layers = 2
|
73 |
+
|
74 |
+
# Parameters for PopArtLayer:
|
75 |
+
# ==============================================================================
|
76 |
+
PopArtLayer.beta = 0.0005
|
77 |
+
PopArtLayer.init_nu = 100.0
|
78 |
+
|
79 |
+
# Parameters for TformerTrajEncoder:
|
80 |
+
# ==============================================================================
|
81 |
+
TformerTrajEncoder.activation = 'leaky_relu'
|
82 |
+
TformerTrajEncoder.causal = True
|
83 |
+
TformerTrajEncoder.d_ff = 5120
|
84 |
+
TformerTrajEncoder.d_model = 1280
|
85 |
+
TformerTrajEncoder.dropout_attn = 0.0
|
86 |
+
TformerTrajEncoder.dropout_emb = 0.05
|
87 |
+
TformerTrajEncoder.dropout_ff = 0.05
|
88 |
+
TformerTrajEncoder.dropout_qkv = 0.0
|
89 |
+
TformerTrajEncoder.head_scaling = True
|
90 |
+
TformerTrajEncoder.n_heads = 20
|
91 |
+
TformerTrajEncoder.n_layers = 9
|
92 |
+
TformerTrajEncoder.norm = 'layer'
|
93 |
+
TformerTrajEncoder.normformer_norms = True
|
94 |
+
TformerTrajEncoder.sigma_reparam = True
|
95 |
+
|
96 |
+
# Parameters for TimestepTransformer:
|
97 |
+
# ==============================================================================
|
98 |
+
# None.
|
99 |
+
|
100 |
+
# Parameters for TokenEmbedding:
|
101 |
+
# ==============================================================================
|
102 |
+
# None.
|
103 |
+
|
104 |
+
# Parameters for TransformerTurnEmbedding:
|
105 |
+
# ==============================================================================
|
106 |
+
TransformerTurnEmbedding.dropout = 0.05
|