hippoformer 0.0.1__tar.gz → 0.0.3__tar.gz

{hippoformer-0.0.1 → hippoformer-0.0.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: hippoformer
-Version: 0.0.1
+Version: 0.0.3
 Summary: hippoformer
 Project-URL: Homepage, https://pypi.org/project/hippoformer/
 Project-URL: Repository, https://github.com/lucidrains/hippoformer

hippoformer-0.0.3/hippoformer/__init__.py

@@ -0,0 +1,4 @@
+from hippoformer.hippoformer import (
+    PathIntegration,
+    mmTEM
+)

hippoformer-0.0.3/hippoformer/hippoformer.py

@@ -0,0 +1,219 @@
+from __future__ import annotations
+
+import torch
+from torch import nn, Tensor, stack, einsum, tensor
+import torch.nn.functional as F
+from torch.nn import Module
+from torch.jit import ScriptModule, script_method
+from torch.func import vmap, grad, functional_call
+
+from einops import repeat, rearrange
+from einops.layers.torch import Rearrange
+
+from x_mlps_pytorch import create_mlp
+
+from assoc_scan import AssocScan
+
+# helpers
+
+def exists(v):
+    return v is not None
+
+def default(v, d):
+    return v if exists(v) else d
+
+def l2norm(t):
+    return F.normalize(t, dim = -1)
+
+# path integration
+
+class RNN(ScriptModule):
+    def __init__(
+        self,
+        dim,
+    ):
+        super().__init__()
+        self.init_hidden = nn.Parameter(torch.randn(1, dim) * 1e-2)
+
+    @script_method
+    def forward(
+        self,
+        transitions: Tensor,
+        hidden: Tensor | None = None
+    ) -> Tensor:
+
+        batch, seq_len = transitions.shape[:2]
+
+        if hidden is None:
+            hidden = l2norm(self.init_hidden)
+            hidden = hidden.expand(batch, -1)
+
+        hiddens: list[Tensor] = []
+
+        for i in range(seq_len):
+            transition = transitions[:, i]
+
+            hidden = einsum('b i, b i j -> b j', hidden, transition)
+            hidden = F.relu(hidden)
+            hidden = l2norm(hidden)
+
+            hiddens.append(hidden)
+
+        return stack(hiddens, dim = 1)
+
+class PathIntegration(Module):
+    def __init__(
+        self,
+        dim_action,
+        dim_structure,
+        mlp_hidden_dim = None,
+        mlp_depth = 2
+    ):
+        # they use the same approach from Ruiqi Gao's paper from 2021
+        super().__init__()
+
+        self.init_structure = nn.Parameter(torch.randn(dim_structure))
+
+        self.to_transitions = create_mlp(
+            default(mlp_hidden_dim,  dim_action * 4),
+            dim_in = dim_action,
+            dim_out = dim_structure * dim_structure,
+            depth = mlp_depth
+        )
+
+        self.mlp_out_to_weights = Rearrange('... (i j) -> ... i j', j = dim_structure)
+
+        self.rnn = RNN(dim_structure)
+
+    def forward(
+        self,
+        actions,                 # (b n d)
+        prev_structural = None   # (b n d) | (b d)
+    ):
+        batch = actions.shape[0]
+
+        transitions = self.to_transitions(actions)
+        transitions = self.mlp_out_to_weights(transitions)
+
+        if exists(prev_structural) and prev_structural.ndim == 3:
+            prev_structural = prev_structural[:, -1]
+
+        return self.rnn(transitions, prev_structural)
+
+# proposed mmTEM
+
+class mmTEM(Module):
+    def __init__(
+        self,
+        dim,
+        *,
+        sensory_encoder: Module,
+        sensory_decoder: Module,
+        dim_sensory,
+        dim_action,
+        dim_encoded_sensory,
+        dim_structure,
+        meta_mlp_depth = 2,
+        decoder_mlp_depth = 2,
+        structure_variance_pred_mlp_depth = 2,
+        path_integrate_kwargs: dict = dict(),
+        loss_weight_generative = 1.,
+        loss_weight_inference = 1.,
+        loss_weight_consistency = 1.,
+        loss_weight_relational = 1.,
+    ):
+        super().__init__()
+
+        # sensory
+
+        self.sensory_encoder = sensory_encoder
+        self.sensory_decoder = sensory_decoder
+
+        dim_joint_rep = dim_encoded_sensory + dim_structure
+
+        self.dim_encoded_sensory = dim_encoded_sensory
+        self.dim_structure = dim_structure
+        self.joint_dims = (dim_structure, dim_encoded_sensory)
+
+        # path integrator
+
+        self.path_integrator = PathIntegration(
+            dim_action = dim_action,
+            dim_structure = dim_structure,
+            **path_integrate_kwargs
+        )
+
+        # meta mlp related
+
+        self.to_queries = nn.Linear(dim_joint_rep, dim, bias = False)
+        self.to_keys = nn.Linear(dim_joint_rep, dim, bias = False)
+        self.to_values = nn.Linear(dim_joint_rep, dim, bias = False)
+
+        self.meta_memory_mlp = create_mlp(
+            dim = dim * 2,
+            depth = meta_mlp_depth,
+            dim_in = dim,
+            dim_out = dim,
+            activation = nn.ReLU()
+        )
+
+        # mlp decoder (from meta mlp output to joint)
+
+        self.memory_output_decoder = create_mlp(
+            dim = dim * 2,
+            dim_in = dim,
+            dim_out = dim_joint_rep,
+            depth = decoder_mlp_depth,
+            activation = nn.ReLU()
+        )
+
+        # the mlp that predicts the variance for the structural code
+        # for correcting the generated structural code modeling the feedback from HC to MEC
+
+        self.structure_variance_pred_mlp_depth = create_mlp(
+            dim = dim_structure * 2,
+            dim_in = dim_structure * 2 + 1,
+            dim_out = dim_structure,
+            depth = structure_variance_pred_mlp_depth
+        )
+
+        # loss related
+
+        self.loss_weight_generative = loss_weight_generative
+        self.loss_weight_inference = loss_weight_inference
+        self.loss_weight_relational = loss_weight_relational
+        self.loss_weight_consistency = loss_weight_consistency
+        self.register_buffer('zero', tensor(0.), persistent = False)
+
+    def forward(
+        self,
+        sensory,
+        actions
+    ):
+        structural_codes = self.path_integrator(actions)
+
+        # first have the structure code be able to fetch from the meta memory mlp
+
+        structure_codes_with_zero_sensory = F.pad(structural_codes, (0, self.dim_encoded_sensory))
+
+        queries = self.to_queries(structure_codes_with_zero_sensory)
+
+        retrieved = self.meta_memory_mlp(queries)
+
+        decoded_structure, decoded_encoded_sensory = self.memory_output_decoder(retrieved).split(self.joint_dims, dim = -1)
+
+        decoded_sensory = self.sensory_decoder(decoded_encoded_sensory)
+
+        generative_pred_loss = F.mse_loss(sensory, decoded_sensory)
+
+        # losses
+
+        total_loss = (
+            generative_pred_loss * self.loss_weight_generative
+        )
+
+        losses = (
+            generative_pred_loss,
+        )
+
+        return total_loss, losses

{hippoformer-0.0.1 → hippoformer-0.0.3}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "hippoformer"
-version = "0.0.1"
+version = "0.0.3"
 description = "hippoformer"
 authors = [
     { name = "Phil Wang", email = "lucidrains@gmail.com" }

hippoformer-0.0.3/tests/test_hippoformer.py

@@ -0,0 +1,37 @@
+import pytest
+
+import torch
+
+def test_path_integrate():
+    from hippoformer.hippoformer import PathIntegration
+
+    path_integrator = PathIntegration(32, 64)
+
+    actions = torch.randn(2, 16, 32)
+
+    structure_codes = path_integrator(actions)
+    structure_codes = path_integrator(actions, structure_codes) # pass in previous structure codes, it will auto use the last one as hidden and pass it to the RNN
+
+    assert structure_codes.shape == (2, 16, 64)
+
+def test_mm_tem():
+    import torch
+    from hippoformer.hippoformer import mmTEM
+
+    from torch.nn import Linear
+
+    model = mmTEM(
+        dim = 32,
+        sensory_encoder = Linear(11, 32),
+        sensory_decoder = Linear(32, 11),
+        dim_sensory = 11,
+        dim_action = 7,
+        dim_structure = 32,
+        dim_encoded_sensory = 32
+    )
+
+    actions = torch.randn(2, 16, 7)
+    sensory = torch.randn(2, 16, 11)
+
+    loss, losses = model(sensory, actions)
+    loss.backward()

hippoformer-0.0.1/hippoformer/hippoformer.py

@@ -1,116 +0,0 @@
-from __future__ import annotations
-
-import torch
-from torch import nn, Tensor, stack, einsum
-import torch.nn.functional as F
-from torch.nn import Module
-from torch.jit import ScriptModule, script_method
-
-from einops import repeat, rearrange
-from einops.layers.torch import Rearrange
-
-from x_mlps_pytorch import create_mlp
-
-from assoc_scan import AssocScan
-
-# helpers
-
-def exists(v):
-    return v is not None
-
-def default(v, d):
-    return v if exists(v) else d
-
-def l2norm(t):
-    return F.normalize(t, dim = -1)
-
-# path integration
-
-class RNN(ScriptModule):
-    def __init__(
-        self,
-        dim,
-    ):
-        super().__init__()
-        self.init_hidden = nn.Parameter(torch.randn(1, dim) * 1e-2)
-
-    @script_method
-    def forward(
-        self,
-        transitions: Tensor,
-        hidden: Tensor | None = None
-    ) -> Tensor:
-
-        batch, seq_len = transitions.shape[:2]
-
-        if hidden is None:
-            hidden = l2norm(self.init_hidden)
-            hidden = hidden.expand(batch, -1)
-
-        hiddens: list[Tensor] = []
-
-        for i in range(seq_len):
-            transition = transitions[:, i]
-
-            hidden = einsum('b i, b i j -> b j', hidden, transition)
-            hidden = F.relu(hidden)
-            hidden = l2norm(hidden)
-
-            hiddens.append(hidden)
-
-        return stack(hiddens, dim = 1)
-
-class PathIntegration(Module):
-    def __init__(
-        self,
-        dim_action,
-        dim_structure,
-        mlp_hidden_dim = None,
-        mlp_depth = 2
-    ):
-        # they use the same approach from Ruiqi Gao's paper from 2021
-        super().__init__()
-
-        self.init_structure = nn.Parameter(torch.randn(dim_structure))
-
-        self.to_transitions = create_mlp(
-            default(mlp_hidden_dim,  dim_action * 4),
-            dim_in = dim_action,
-            dim_out = dim_structure * dim_structure,
-            depth = mlp_depth
-        )
-
-        self.mlp_out_to_weights = Rearrange('... (i j) -> ... i j', j = dim_structure)
-
-        self.rnn = RNN(dim_structure)
-
-    def forward(
-        self,
-        actions,                 # (b n d)
-        prev_structural = None   # (b n d) | (b d)
-    ):
-        batch = actions.shape[0]
-
-        transitions = self.to_transitions(actions)
-        transitions = self.mlp_out_to_weights(transitions)
-
-        if exists(prev_structural) and prev_structural.ndim == 3:
-            prev_structural = prev_structural[:, -1]
-
-        return self.rnn(transitions, prev_structural)
-
-# proposed mmTEM
-
-class mmTEM(Module):
-    def __init__(
-        self,
-        dim
-    ):
-        super().__init__()
-
-
-    def forward(
-        self,
-        data
-    ):
-        raise NotImplementedError

hippoformer-0.0.1/tests/test_hippoformer.py

@@ -1,15 +0,0 @@
-import pytest
-
-import torch
-
-def test_path_integrate():
-    from hippoformer.hippoformer import PathIntegration
-
-    path_integrator = PathIntegration(32, 64)
-
-    actions = torch.randn(2, 16, 32)
-
-    structure_codes = path_integrator(actions)
-    structure_codes = path_integrator(actions, structure_codes) # pass in previous structure codes, it will auto use the last one as hidden and pass it to the RNN
-
-    assert structure_codes.shape == (2, 16, 64)