Upload folder using huggingface_hub

small-rl-maxq/ckpts/policy_weights/policy_epoch_0.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:239def3b32a1244ae00b5978d43043ba4f8d65a107d8f4e7e0d52e67c145a03a
+size 55821590

small-rl-maxq/ckpts/policy_weights/policy_epoch_10.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3b2bc80e4b6f2bffefc4be2299e7b3234841b45f54ecb2edce439d85afa2cba7
+size 55821762

small-rl-maxq/ckpts/policy_weights/policy_epoch_12.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4222369b416d164816491595d159a7943468637fad684eab6031a57efec27e54
+size 55821762

small-rl-maxq/ckpts/policy_weights/policy_epoch_14.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6990cd796fe57ae104353e64ba88b5edb7f9fbc55c553c058d8b033f06da2ece
+size 55821762

small-rl-maxq/ckpts/policy_weights/policy_epoch_16.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:882bbcf1de6e60e14cf0175940779f8e9af40afa5eb8c33ffb28516eb916184e
+size 55821762

small-rl-maxq/ckpts/policy_weights/policy_epoch_18.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f2bc15729389cbeedbe12f5e8c06a74f5eed667b5654fd3b764dae9facc7e8ec
+size 55821762

small-rl-maxq/ckpts/policy_weights/policy_epoch_2.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c1ceb34b42b3dd4443ffe47aa672680ab30983a4de4e3013fe144f7abea890cd
+size 55821590

small-rl-maxq/ckpts/policy_weights/policy_epoch_20.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4afd037f2e35f0bea55ee3796a49749ebe83fd96193a108c618df497346e2a06
+size 55821762

small-rl-maxq/ckpts/policy_weights/policy_epoch_22.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b2e708ca1d747e87ae51719fae326c9b034ee078d174b4260dcbacaf52a0fc21
+size 55821762

small-rl-maxq/ckpts/policy_weights/policy_epoch_24.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bafaff29549e42885d4eaa14a0c6419b7eff51a44454439846fd1370e1c9e7d9
+size 55821762

small-rl-maxq/ckpts/policy_weights/policy_epoch_26.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:11cfcee49602101f2bfc8768da7f1c8164afece74b2bd3a60df1db1274c73574
+size 55821762

small-rl-maxq/ckpts/policy_weights/policy_epoch_28.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b5812ce2f04a83e0d9945446ca4f05c2ad56e3e726ccd53b09c0648c0018030e
+size 55821762

small-rl-maxq/ckpts/policy_weights/policy_epoch_30.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:35677cd53793020e9a1988c2290473aeadadea5b3acb8eed22ad14cdb211e0bb
+size 55821762

small-rl-maxq/ckpts/policy_weights/policy_epoch_32.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:32d7907f4eb47e02ff944ff04a89e3940ce1b44c3a7a6ef1d64973b8a17f4fd0
+size 55821762

small-rl-maxq/ckpts/policy_weights/policy_epoch_34.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d07dd978f2e8da2ed812e91870b0c68982f22205465d0da945349039d87e9b3a
+size 55821762

small-rl-maxq/ckpts/policy_weights/policy_epoch_36.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ec10ea88e6ca3114a2604570edc23f57f663f9b128c443a38ff8b6acd899424d
+size 55821762

small-rl-maxq/ckpts/policy_weights/policy_epoch_38.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3d289cd8b968ee4fdd864dcecd8443f5bd95d6ca3e3934f38c69472308dae358
+size 55821762

small-rl-maxq/ckpts/policy_weights/policy_epoch_4.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bf83fc62dc6ff8865d220f4fa2ee0a2d9f5eb0fdaa91ff448f95195ed32a385d
+size 55821590

small-rl-maxq/ckpts/policy_weights/policy_epoch_40.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:82d02dcd00da93268278bab2f7bbfccdd1e681017194618cd937dd4c1fafe071
+size 55821762

small-rl-maxq/ckpts/policy_weights/policy_epoch_42.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f2479f5f121053c3f7ca195782e2b8405cd6902db7dc0bf5b80b6375d806ccc7
+size 55821762

small-rl-maxq/ckpts/policy_weights/policy_epoch_44.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e79375b39f1e557ad12fc5dc5921608a886a37e554552d07e7678119025fb397
+size 55821762

small-rl-maxq/ckpts/policy_weights/policy_epoch_46.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1a70e7762ec263b7cb19056001dcf03976b54c59e1b8a38a1dce4b4078c7b6c2
+size 55821762

small-rl-maxq/ckpts/policy_weights/policy_epoch_48.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2c34227f55b81360d1a91f4dee11db86416c7bfbb4fbb0edfd5f46b5424ab4d7
+size 55821762

small-rl-maxq/ckpts/policy_weights/policy_epoch_50.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eb0af867e34b189ebe1df07154381aef1517ff434547b9eb67601654648be2c9
+size 55821762

small-rl-maxq/ckpts/policy_weights/policy_epoch_52.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:77f2b620d2f58c4bf777e9f6fc57277360eb120a49e0a8f6b8f95b00cbfe7260
+size 55821762

small-rl-maxq/ckpts/policy_weights/policy_epoch_54.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:77d5766314a38884770184c0873ce2b7c8048bcc76d64013f29ae64d1e0e840c
+size 55821762

small-rl-maxq/ckpts/policy_weights/policy_epoch_6.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fb0035b5caa90541a125c4cafa141b061b1eb032ae60f36a8daee650aaca4982
+size 55821590

small-rl-maxq/ckpts/policy_weights/policy_epoch_8.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6aa75f7569835955ff0ac37d982e100a7f396e9bebfc5c2cd6db4b3e6977870a
+size 55821590

small-rl-maxq/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