A head on an encoder's representations (last_hidden_state), for fine-tuning or probing. Pooling::Mean averages each sequence's residues (not <cls>, <eos> or padding) and applies a linear layer: (N, L, hidden) to (N, outputs), for per-protein labels. Pooling::PerResidue applies it to every token: (N, L, hidden) to (N, L, outputs), for per-residue labels (use -100 targets at <cls>, <eos> and padding).
More...
#include <protein_training.hpp>
|
| | SequenceHead (int64_t hidden_size, int64_t outputs, Pooling pooling, DeviceBackend *backend, uint64_t seed=0) |
| | The linear layer starts as transformers' heads do: weights from N(0, 0.02²), drawn from seed, and zero bias.
|
| |
| void | set_residue_mask (const Tensor &residues) |
| | Which tokens Pooling::Mean averages: (N, L), 1 for a residue. Set it before every forward() whose shape changes; ResidueMask() builds it.
|
| |
| Tensor | backward (const Tensor &grad_output) override |
| | Computes the gradient w.r.t. this module's input, given the gradient w.r.t. its output. Must be called after forward() – uses state cached from that call.
|
| |
| Tensor | propagate_relevance (const Tensor &relevance_out, const LRPRuleConfig &config) override |
| | Relevance at the representations, (N, L, hidden): the linear layer's rule, then (for Mean) each residue's share of the average.
|
| |
| OpType | op_type () const override |
| | This module's operation-category tag, for ComputationGraph node tagging.
|
| |
| std::vector< NamedParamRef > | named_parameters () override |
| | This module's trainable parameters, each with its hierarchical name – the one place a module declares its parameters (roadmap FND-1).
|
| |
| void | set_training (bool training) override |
| | Sets this module's training/eval mode. Defaults to training (matches every mainstream framework's Module default).
|
| |
| std::optional< DeviceType > | compute_device () const override |
| | The device this module computes on, so forward() can reject an input on another device before any kernel sees it (roadmap FND-8).
|
| |
| void | release_activations () override |
| | Frees the tensors forward() kept for backward() and the relevance passes; containers release their children too. Inference that only needs the output calls it after each forward, so a deep model holds one layer's activations at a time (a 1024-token ESM-2 650M pass otherwise keeps over 10 GB of them).
|
| |
| LinearModule & | linear () |
| |
| Pooling | pooling () const |
| |
| virtual | ~Module ()=default |
| |
| Tensor | forward (const Tensor &input) |
| | Runs this module's forward computation.
|
| |
| std::pair< Tensor, NodeId > | forward_traced (const Tensor &input, NodeId input_node, ComputationGraph &graph, Autograd &autograd) |
| | Runs forward() while also registering a ComputationGraph node (tagged with this module's op_type(), parented to input_node) and wiring an Autograd backward function that reuses this module's own backward() – the opt-in traced/explainable path, per Phase 2 Mission 0.
|
| |
| virtual bool | supports_lrp_rule (LRPRule rule) const |
| | Whether propagate_relevance() implements rule (no silent fallback: callers such as ExplainerContext::relevance_pass() throw rather than run a module on a rule it does not implement).
|
| |
| virtual std::vector< NamedBufferRef > | named_buffers () |
| | This module's buffers: state a checkpoint must save that no optimizer updates.
|
| |
| virtual std::vector< NamedModuleRef > | named_children () |
| | This module's direct sub-modules, each under the prefix its parameters carry in named_parameters(), in the same order (TRN-7: how ApplyLora() finds a model's Linear layers). Default: none. NamedModules() walks the whole tree.
|
| |
| virtual std::vector< ParamRef > | parameters () |
| | This module's trainable parameters and their gradients, for an optimizer to update uniformly across module types.
|
| |
| void | set_requires_grad (bool requires_grad, const std::string &prefix="") |
| | Freezes (false) or unfreezes (true) parameters by name (roadmap FND-2).
|
| |
| bool | is_training () const |
| | Whether this module is currently in training mode.
|
| |
A head on an encoder's representations (last_hidden_state), for fine-tuning or probing. Pooling::Mean averages each sequence's residues (not <cls>, <eos> or padding) and applies a linear layer: (N, L, hidden) to (N, outputs), for per-protein labels. Pooling::PerResidue applies it to every token: (N, L, hidden) to (N, L, outputs), for per-residue labels (use -100 targets at <cls>, <eos> and padding).
Train it alone on a frozen encoder (a probe), or with it: pass backward()'s result to EncoderLM::backward_hidden(). Parameters are named linear.weight and linear.bias.
◆ Pooling
| Enumerator |
|---|
| Mean | |
| PerResidue | |
◆ SequenceHead()
| pulsatrix::SequenceHead::SequenceHead |
( |
int64_t |
hidden_size, |
|
|
int64_t |
outputs, |
|
|
Pooling |
pooling, |
|
|
DeviceBackend * |
backend, |
|
|
uint64_t |
seed = 0 |
|
) |
| |
The linear layer starts as transformers' heads do: weights from N(0, 0.02²), drawn from seed, and zero bias.
◆ backward()
| Tensor pulsatrix::SequenceHead::backward |
( |
const Tensor & |
grad_output | ) |
|
|
overridevirtual |
Computes the gradient w.r.t. this module's input, given the gradient w.r.t. its output. Must be called after forward() – uses state cached from that call.
- Parameters
-
| grad_output | Gradient w.r.t. this module's output. |
- Returns
- Gradient w.r.t. this module's input.
- Note
- Promoted to the base class in Phase 2 Mission 0 – every existing subclass (LinearModule/ReluModule/Conv2DModule) already implemented this exact signature independently; making it virtual lets graph-wiring code (Module::forward_traced) call it polymorphically through a Module* without knowing the concrete subclass, the same way propagate_relevance already works.
Implements pulsatrix::Module.
◆ compute_device()
| std::optional< DeviceType > pulsatrix::SequenceHead::compute_device |
( |
| ) |
const |
|
inlineoverridevirtual |
The device this module computes on, so forward() can reject an input on another device before any kernel sees it (roadmap FND-8).
- Returns
- std::nullopt (the default) skips the check: a container whose layers check their own inputs, or a user module written before this existed. Every in-tree layer returns its device; EmbeddingModule doesn't, because it reads its indices through their own backend and so accepts them from any device.
Reimplemented from pulsatrix::Module.
◆ forward_impl()
| Tensor pulsatrix::SequenceHead::forward_impl |
( |
const Tensor & |
input | ) |
|
|
overrideprotectedvirtual |
- Exceptions
-
| std::invalid_argument | unless the input is (N, L, hidden), or for Mean, the residue mask is set and matches (N, L). |
Implements pulsatrix::Module.
◆ linear()
◆ named_parameters()
| std::vector< NamedParamRef > pulsatrix::SequenceHead::named_parameters |
( |
| ) |
|
|
overridevirtual |
This module's trainable parameters, each with its hierarchical name – the one place a module declares its parameters (roadmap FND-1).
- Returns
- {name, {value, grad}} entries pointing directly at this module's own members, in a fixed order. Names are unique within the module tree. Default: empty (a parameterless module like ReluModule needs no override).
- Note
- Override this, not parameters(): saving, loading, freezing by name and optimizer parameter groups all key on these names.
Reimplemented from pulsatrix::Module.
◆ op_type()
| OpType pulsatrix::SequenceHead::op_type |
( |
| ) |
const |
|
inlineoverridevirtual |
This module's operation-category tag, for ComputationGraph node tagging.
- Returns
- This module's OpType (see op_type.hpp's closed set).
- Note
- Added in Phase 2 Mission 0 alongside backward() – lets graph-wiring code tag nodes generically through a Module* rather than switching on concrete subclass.
Implements pulsatrix::Module.
◆ pooling()
| Pooling pulsatrix::SequenceHead::pooling |
( |
| ) |
const |
|
inline |
◆ propagate_relevance()
Relevance at the representations, (N, L, hidden): the linear layer's rule, then (for Mean) each residue's share of the average.
Implements pulsatrix::Module.
◆ release_activations()
| void pulsatrix::SequenceHead::release_activations |
( |
| ) |
|
|
overridevirtual |
Frees the tensors forward() kept for backward() and the relevance passes; containers release their children too. Inference that only needs the output calls it after each forward, so a deep model holds one layer's activations at a time (a 1024-token ESM-2 650M pass otherwise keeps over 10 GB of them).
- Note
- Until the next forward(), backward() and relevance calls are invalid. The default has nothing to free.
Reimplemented from pulsatrix::Module.
◆ set_residue_mask()
| void pulsatrix::SequenceHead::set_residue_mask |
( |
const Tensor & |
residues | ) |
|
◆ set_training()
| void pulsatrix::SequenceHead::set_training |
( |
bool |
training | ) |
|
|
overridevirtual |
Sets this module's training/eval mode. Defaults to training (matches every mainstream framework's Module default).
- Note
- Virtual since Phase 6 Mission 6 (SequentialModule) – Mission 5 originally shipped this as plain non-virtual state ("no composite container exists yet to
cascade through"); SequentialModule overrides this to cascade to every contained layer, and needs virtual dispatch to do so correctly even when accessed through a Module* base pointer, not just its own concrete type.
-
BatchNormModule keeps running statistics in training mode and normalizes with them in eval mode; DropoutModule is the identity in eval mode. Call set_training(false) before explaining a model that has either, so a sample's explanation doesn't depend on the rest of its batch.
Reimplemented from pulsatrix::Module.
The documentation for this class was generated from the following file: