Upload folder using huggingface_hub

synthetic-rl-v1++/ckpts/policy_weights/policy_epoch_28.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ae0348270edabea5fa904ee97b16bcf5e68b3415745e32c9b9a92e7dd5f88bd3
+size 802136514

synthetic-rl-v1++/ckpts/policy_weights/policy_epoch_29.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c8364b33cb91a41a25b1ce6907ed46d1d0b24d9b48d0482241b6c44378fc0037
+size 802136514

synthetic-rl-v1++/ckpts/policy_weights/policy_epoch_30.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3ca370aa12e184b90d240d57a9355888559317fd654fddf4b2c4c619c98299eb
+size 802136514

synthetic-rl-v1++/ckpts/policy_weights/policy_epoch_31.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:12fa234dbf6ba14f0777bfc5f529dd430d999b137be9ecbfd27fb4bd6bb39d60
+size 802136514

synthetic-rl-v1++/ckpts/policy_weights/policy_epoch_32.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:59433f831c820f220cd8e5dfa17ca5486e58ff830005245ed891443e63c71483
+size 802136514

synthetic-rl-v1++/ckpts/policy_weights/policy_epoch_33.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c7a1ff6fda63baaaf14e8a671bb36a2e850da8969ec5db01285685989b3816af
+size 802136514

synthetic-rl-v1++/ckpts/policy_weights/policy_epoch_34.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ea3666530236ca62a6829263a9ed339928dc009369bc09c2a30d9225f24a0784
+size 802136514

synthetic-rl-v1++/ckpts/policy_weights/policy_epoch_35.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5996d3c30a4fc81b63ed741e21f81acf0cfd84d91c78fe537ee54eb9c788379f
+size 802136514

synthetic-rl-v1++/ckpts/policy_weights/policy_epoch_36.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a52e3300aece457daeb1112ad6a54eac05ca44847e36d9a929a53ddf26971ffa
+size 802136514

synthetic-rl-v1++/ckpts/policy_weights/policy_epoch_37.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:896b135450fd31c83d05a205ed3cdc9412c113b239d300c238ef4b7cfef52047
+size 802136514

synthetic-rl-v1++/ckpts/policy_weights/policy_epoch_38.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e335dea6f98cf4f18dab8c1c0dc212d35cef923e52ce6364373eea3ac747613b
+size 802136514

synthetic-rl-v1++/ckpts/policy_weights/policy_epoch_39.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:745aba18669c31cf86f6c7e373251edf7b1a3b6f35cb3f85c580cc624086975b
+size 802136514

synthetic-rl-v1++/config.txt

@@ -0,0 +1,106 @@
+# Parameters for Actor:
+# ==============================================================================
+Actor.activation = 'leaky_relu'
+Actor.cont_dist_kind = 'normal'
+Actor.d_hidden = 512
+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 = 6
+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 = 160
+MetamonTstepEncoder.extra_emb_dim = 18
+MetamonTstepEncoder.n_heads = 8
+MetamonTstepEncoder.n_layers = 5
+MetamonTstepEncoder.scratch_tokens = 11
+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 = 512
+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 = 5120
+TformerTrajEncoder.d_model = 1280
+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 = 20
+TformerTrajEncoder.n_layers = 9
+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