Upload folder using huggingface_hub

small-rl/ckpts/policy_weights/policy_epoch_0.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bb96980066b5d476aa43adc451e0465e545aa87447ed0a88244fc28dfa469f83
+size 55821590

small-rl/ckpts/policy_weights/policy_epoch_10.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e4cdcb603cfd8c89cf72bea9ebcc39dda550a167ed0509935e76ee1e3e27e36b
+size 55821762

small-rl/ckpts/policy_weights/policy_epoch_12.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:491a82b1a2ab590c99b60b8a17824d227d75b141cc35be4740b025577329b0ff
+size 55821762

small-rl/ckpts/policy_weights/policy_epoch_14.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:73df41842637b7fe3b4614d42f1f134cdc1718da591d1283b1b1b213da7b2dc2
+size 55821762

small-rl/ckpts/policy_weights/policy_epoch_16.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:51f8d29b83927e225ce0f150f15b50ee2dcb0a529772c4511dcbddcbd3f315bd
+size 55821762

small-rl/ckpts/policy_weights/policy_epoch_18.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8fd551dc1bb49808518e21b8e866cf8b8943129b20a413a83f14dd1c10b7e7d3
+size 55821762

small-rl/ckpts/policy_weights/policy_epoch_2.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d1cdf343d8d983ab64e755f5f7baba13e2f6535032c79831f297bdae27f3ff40
+size 55821590

small-rl/ckpts/policy_weights/policy_epoch_20.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:69c2a376bffc090ca171d84834d257b6f65b95fcf7c907a9d527e8ae4f8ff971
+size 55821762

small-rl/ckpts/policy_weights/policy_epoch_22.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d699d09e3db02cc9cddd84cbbe21a0843d2b9aad0fb8829b22e1f792b2e700dc
+size 55821762

small-rl/ckpts/policy_weights/policy_epoch_24.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cd80e692a4a6a87c5c6eb157e6f4b1dfa0c5bbe8c15bd994629fd704ff07d8d6
+size 55821762

small-rl/ckpts/policy_weights/policy_epoch_26.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:82cd658276c2e57b12f799e8a684a609c3ef4744505f8eed568f062ec7852d49
+size 55821762

small-rl/ckpts/policy_weights/policy_epoch_28.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:38a5167462c81493b96c865e3b0ff28cb6c5e610cfcf9101923b3ea9bfba767d
+size 55821762

small-rl/ckpts/policy_weights/policy_epoch_30.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:262dd25bb5d401cae93de2e663db582bb63757ff3f74837a5fc8f6622fb18133
+size 55821762

small-rl/ckpts/policy_weights/policy_epoch_32.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b23f2d10757740b32c78b2c2ac518233b84dad49563e2ed5910b492db12b65dc
+size 55821762

small-rl/ckpts/policy_weights/policy_epoch_34.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b4b244c65034a00c25da2051f219161fe1a69efb1c2be711024b2c28c7843973
+size 55821762

small-rl/ckpts/policy_weights/policy_epoch_36.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dd0dd6760ce78d1ff57d2f3461e63f2049ada9439ecd3cfe49c49c4020060230
+size 55821762

small-rl/ckpts/policy_weights/policy_epoch_38.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2b8ab58cdc6b91f1fa91e56ae227419f460924a56c89d98016ffc4ebcda73d2a
+size 55821762

small-rl/ckpts/policy_weights/policy_epoch_4.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:061da09016c075ae2571a54707fb2b44dc09fe4c5287936f17be75a85922383f
+size 55821590

small-rl/ckpts/policy_weights/policy_epoch_40.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f4258211777480643fbfe573ff906a2ef1a86b0bc2ad9cce117efdc797e40369
+size 55821762

small-rl/ckpts/policy_weights/policy_epoch_6.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f1c397a99c1a33fdb0dde778ea1a4ced83f9003eed9523a7dce1a553d87868e5
+size 55821590

small-rl/ckpts/policy_weights/policy_epoch_8.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5391debf5b27dc084beae69ada0d8b35c061efef9d71b955da934fcc186df6a9
+size 55821590

small-rl/config.txt

@@ -0,0 +1,106 @@
+# Parameters for Actor:
+# ==============================================================================
+Actor.activation = 'leaky_relu'
+Actor.cont_dist_kind = 'normal'
+Actor.d_hidden = 300
+Actor.dropout_p = 0.0
+Actor.gmm_modes = 5
+Actor.log_std_high = 2.0
+Actor.log_std_low = -5.0
+Actor.n_layers = 2
+
+# Parameters for Agent:
+# ==============================================================================
+Agent.fake_filter = False
+Agent.gamma = 0.999
+Agent.num_critics = 4
+Agent.num_critics_td = 2
+Agent.offline_coeff = 1.0
+Agent.online_coeff = 0.0
+Agent.popart = True
+Agent.reward_multiplier = 10.0
+Agent.tau = 0.003
+Agent.use_multigamma = True
+Agent.use_target_actor = True
+
+# Parameters for Experiment:
+# ==============================================================================
+Experiment.batches_per_update = 1
+Experiment.critic_loss_weight = 10.0
+Experiment.env_mode = 'async'
+Experiment.force_reset_train_envs_every = None
+Experiment.grad_clip = 1.0
+Experiment.has_replay_buffer_rights = True
+Experiment.l2_coeff = 0.001
+Experiment.learning_rate = 0.0001
+Experiment.local_time_optimizer = False
+Experiment.lr_warmup_steps = 500
+Experiment.mixed_precision = 'no'
+Experiment.padded_sampling = 'none'
+Experiment.save_trajs_as = 'npz'
+Experiment.stagger_traj_file_lengths = True
+Experiment.wandb_group_name = None
+
+# Parameters for FlashAttention:
+# ==============================================================================
+FlashAttention.window_size = (-1, -1)
+
+# Parameters for MetamonTstepEncoder:
+# ==============================================================================
+MetamonTstepEncoder.d_model = 100
+MetamonTstepEncoder.extra_emb_dim = 18
+MetamonTstepEncoder.n_heads = 5
+MetamonTstepEncoder.n_layers = 3
+MetamonTstepEncoder.scratch_tokens = 4
+MetamonTstepEncoder.token_mask_aug = False
+
+# Parameters for Multigammas:
+# ==============================================================================
+Multigammas.continuous = [0.1, 0.9, 0.95, 0.97, 0.99, 0.995]
+Multigammas.discrete = [0.1, 0.9, 0.95, 0.97, 0.99, 0.995]
+
+# Parameters for MultiModalEmbedding:
+# ==============================================================================
+MultiModalEmbedding.dropout = 0.05
+MultiModalEmbedding.numerical_tokens = 6
+
+# Parameters for NCritics:
+# ==============================================================================
+NCritics.activation = 'leaky_relu'
+NCritics.d_hidden = 300
+NCritics.dropout_p = 0.0
+NCritics.n_layers = 2
+
+# Parameters for PopArtLayer:
+# ==============================================================================
+PopArtLayer.beta = 0.0005
+PopArtLayer.init_nu = 100.0
+
+# Parameters for TformerTrajEncoder:
+# ==============================================================================
+TformerTrajEncoder.activation = 'leaky_relu'
+TformerTrajEncoder.causal = True
+TformerTrajEncoder.d_ff = 2048
+TformerTrajEncoder.d_model = 512
+TformerTrajEncoder.dropout_attn = 0.0
+TformerTrajEncoder.dropout_emb = 0.05
+TformerTrajEncoder.dropout_ff = 0.05
+TformerTrajEncoder.dropout_qkv = 0.0
+TformerTrajEncoder.head_scaling = True
+TformerTrajEncoder.n_heads = 8
+TformerTrajEncoder.n_layers = 3
+TformerTrajEncoder.norm = 'layer'
+TformerTrajEncoder.normformer_norms = True
+TformerTrajEncoder.sigma_reparam = True
+
+# Parameters for TimestepTransformer:
+# ==============================================================================
+# None.
+
+# Parameters for TokenEmbedding:
+# ==============================================================================
+# None.
+
+# Parameters for TransformerTurnEmbedding:
+# ==============================================================================
+TransformerTurnEmbedding.dropout = 0.05