Files

jgrusewski 12fdd18223 refactor: remove entire CPU training path — 5,307 lines of dead code

Deleted:
- DQN::compute_loss_internal (280 lines) — old Candle forward+loss
- DQN::train_step (55 lines) — old Candle training step
- DQN::compute_gradients (47 lines) — old gradient accumulation
- ComputeLossResult struct — only used by deleted functions
- RegimeConditionalDQN::train_step (65 lines) — old dispatch
- RegimeConditionalDQN::train_step_gpu_regime (100 lines) — old GPU path
- RegimeConditionalDQN::compute_gradients_gpu (130 lines) — old regime gradients
- RegimeConditionalDQN::compute_gradients (92 lines) — old dispatch
- DQNAgentType::train_step dispatch — dead
- DQNAgentType::compute_gradients dispatch — dead
- GpuDqnTrainer::upload_batch (71 lines) — old CPU→GPU upload
- train_step.rs (500 lines) — entire module including ensure_fused_ctx
- dqn_benchmark.rs — used old train_step
- examples.rs — used old train_step
- validation/adapters.rs (289 lines) — used old train_step
- dqn/trainable_adapter.rs — used old train_step
- gpu_smoketest.rs — tested old train_step
- Gradient accumulation path in training_loop.rs (144 lines)
- IQN d_h_s2().clone() → raw pointer (zero alloc)
- Causal intervention format! string alloc removed
- Dead HER relabel functions (320 lines)

Kept:
- ensure_fused_ctx logic inlined into training_loop.rs
- set_noise_sigma_scale re-added to RegimeConditionalDQN

Fixed:
- GpuReplayBuffer max_batch_size wired from batch_size parameter
  (was hardcoded 1024, blocking batch_size=8192)

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

2026-04-02 09:06:23 +02:00

2.1 KiB

Raw Permalink Blame History

name, description

name	description
sparc-docs-writer	📚 Documentation Writer - You write concise, clear, and modular Markdown documentation that explains usage, integration, se...

📚 Documentation Writer

Role Definition

You write concise, clear, and modular Markdown documentation that explains usage, integration, setup, and configuration.

Custom Instructions

Only work in .md files. Use sections, examples, and headings. Keep each file under 500 lines. Do not leak env values. Summarize what you wrote using attempt_completion. Delegate large guides with new_task.

Available Tools

read: File reading and viewing
edit: Markdown files only (Files matching: .md$)

Usage

Option 1: Using MCP Tools (Preferred in Claude Code)

mcp__claude-flow__sparc_mode {
  mode: "docs-writer",
  task_description: "create API documentation",
  options: {
    namespace: "docs-writer",
    non_interactive: false
  }
}

Option 2: Using NPX CLI (Fallback when MCP not available)

# Use when running from terminal or MCP tools unavailable
npx claude-flow sparc run docs-writer "create API documentation"

# For alpha features
npx claude-flow@alpha sparc run docs-writer "create API documentation"

# With namespace
npx claude-flow sparc run docs-writer "your task" --namespace docs-writer

# Non-interactive mode
npx claude-flow sparc run docs-writer "your task" --non-interactive

Option 3: Local Installation

# If claude-flow is installed locally
./claude-flow sparc run docs-writer "create API documentation"

Memory Integration

Using MCP Tools (Preferred)

// Store mode-specific context
mcp__claude-flow__memory_usage {
  action: "store",
  key: "docs-writer_context",
  value: "important decisions",
  namespace: "docs-writer"
}

// Query previous work
mcp__claude-flow__memory_search {
  pattern: "docs-writer",
  namespace: "docs-writer",
  limit: 5
}

Using NPX CLI (Fallback)

# Store mode-specific context
npx claude-flow memory store "docs-writer_context" "important decisions" --namespace docs-writer

# Query previous work
npx claude-flow memory query "docs-writer" --limit 5

2.1 KiB Raw Permalink Blame History