Yif-Yang
106 lines
4.2 KiB
Markdown
106 lines
4.2 KiB
Markdown
# The Training Loop
|
|
|
|
SkillOpt's core insight: **optimizing natural-language skill documents follows the same structure as training neural networks**.
|
|
|
|
## Overview
|
|
|
|
```
|
|
┌─────────────────────────────────────────────────────────┐
|
|
│ Training Loop │
|
|
│ │
|
|
│ for epoch in epochs: │
|
|
│ for step in steps: │
|
|
│ 1. Rollout — Target executes tasks │
|
|
│ 2. Reflect — Optimizer analyzes trajectories │
|
|
│ 3. Aggregate — Hierarchical merge of patches │
|
|
│ 4. Select — Rank & clip edits (learning rate) │
|
|
│ 5. Update — Apply patches to skill doc │
|
|
│ 6. Gate — Validate & accept/reject │
|
|
│ │
|
|
│ Epoch Boundary: │
|
|
│ • Slow Update (longitudinal comparison & guidance) │
|
|
│ • Meta Skill (cross-epoch strategy memory) │
|
|
└─────────────────────────────────────────────────────────┘
|
|
```
|
|
|
|
## Stage Details
|
|
|
|
### 1. Rollout (Forward Pass)
|
|
|
|
The **target** model executes tasks using the current skill document as its prompt. Each task produces a trajectory and a score.
|
|
|
|
```python
|
|
# Analogy: forward pass through the network
|
|
predictions = model(input, skill_document)
|
|
scores = evaluate(predictions, ground_truth)
|
|
```
|
|
|
|
### 2. Reflect (Backward Pass)
|
|
|
|
The **optimizer** model analyzes trajectory minibatches and produces **edit
|
|
patches** — structured suggestions for improving the skill document. Failure
|
|
minibatches are always eligible for analysis; successful trajectories are also
|
|
analyzed unless `gradient.failure_only` is enabled. Independent minibatches can
|
|
run concurrently according to `gradient.analyst_workers`.
|
|
|
|
```python
|
|
# Analogy: computing gradients
|
|
gradients = loss.backward() # → edit patches
|
|
```
|
|
|
|
### 3. Aggregate
|
|
|
|
Semantically similar edit patches are merged to avoid redundant edits.
|
|
|
|
### 4. Select (Gradient Clipping)
|
|
|
|
Edits are ranked by relevance score. The `learning_rate` parameter caps how many edits are applied per step — just like gradient clipping prevents overshooting.
|
|
|
|
```python
|
|
# Analogy: gradient clipping + optimizer step size
|
|
selected = top_k(edits, k=learning_rate)
|
|
```
|
|
|
|
The `lr_scheduler` adjusts this over training:
|
|
|
|
- **cosine**: Start aggressive, taper smoothly
|
|
- **linear**: Linear decay
|
|
- **constant**: Fixed rate
|
|
|
|
### 5. Update (Parameter Update)
|
|
|
|
Selected edits are applied to the skill document, producing a new version.
|
|
|
|
### 6. Gate (Validation)
|
|
|
|
The updated skill is evaluated on a **selection split** (analogous to a
|
|
validation set). With the gate enabled, the candidate is accepted only when its
|
|
configured gate score (`hard`, `soft`, or `mixed`) is strictly higher than the
|
|
current skill's score. With `evaluation.use_gate: false`, validation is still
|
|
recorded but candidates are force-accepted.
|
|
|
|
## Epoch Boundary Mechanisms
|
|
|
|
### Slow Update
|
|
|
|
At the end of each epoch (starting from epoch 2), the system performs a
|
|
**longitudinal comparison**: it rolls out both the previous epoch's skill and
|
|
the current skill on the same samples, categorizes items as
|
|
improved/regressed/persistent-fail/stable-success, then generates high-level
|
|
**guidance** for the skill document. Depending on
|
|
`optimizer.slow_update_gate_with_selection`, that guidance is either checked on
|
|
the selection split or applied unconditionally. Its purpose is to counter
|
|
cross-epoch forgetting.
|
|
|
|
### Meta Skill
|
|
|
|
A **meta-skill memory** accumulates high-level strategy notes across the training
|
|
run. Starting at the end of epoch 2, the optimizer compares the previous and
|
|
current epoch, writes a compact memory, and provides the prior epoch's memory as
|
|
additional context during later reflection and update stages.
|
|
|
|
## Next Steps
|
|
|
|
- [Understand Skill Documents](skill-document.md)
|
|
- [DL ↔ SkillOpt analogy table](dl-analogy.md)
|