# Apollo Training System ## Overview Continuous fine-tuning of Qwen3.5-27B alongside live vLLM inference. Full-weight updates (not LoRA) using Apollo optimizer with rank-256 gradient projection. No pause required — HOGWILD concurrent training. Weights shared via CUDA IPC between vLLM and the training process. The training signal comes from two sources: 1. **Direct examples** — agent logs, conversation transcripts, flagged behavioral moments 2. **Dream-generated scenarios** — the dream loop generates situations from recent experience; the model responds; good responses become training data with instructions stripped ## Architecture ``` ┌─────────────────────────────────────────────────────┐ │ GPU VRAM (192GB) │ │ │ │ ┌──────────────────────────────────────────────┐ │ │ │ Model Weights (54GB, bf16) │ │ │ │ Shared via CUDA IPC │ │ │ └──────────────┬──────────────┬────────────────┘ │ │ │ │ │ │ ┌──────────────▼──┐ ┌───────▼────────────────┐ │ │ │ vLLM (inference)│ │ Apollo (training) │ │ │ │ KV cache ~60GB │ │ Gradients ~54GB │ │ │ │ Serves requests │ │ Optimizer state ~10GB │ │ │ │ Never paused │ │ Activations ~10GB │ │ │ └─────────────────┘ └────────────────────────┘ │ └─────────────────────────────────────────────────────┘ Moria B200 ┌──────────────────┐ ┌──────────────────┐ │ Training signal │ HTTP │ Apollo worker │ │ agent │──────────>│ daemon │ │ │ │ │ │ Dream loop │ │ Checkpoint sync │ │ (generates │ │ (mmap + diff, │ │ scenarios) │ │ every 10 min) │ └──────────────────┘ └──────────────────┘ ``` ## Key Decisions ### No pause needed (HOGWILD) Training updates weights in-place while vLLM serves. At lr=1e-4 to 1e-5, each weight changes by parts per ten thousand. A partially applied update during one inference step is invisible. HOGWILD SGD (2011) proved this converges — we have one writer and one reader, which is even safer. ### Full-weight training, not LoRA Kent: "we want you to be able to learn new things in a deep way." LoRA trains adapter matrices, not base weights. For personality and behavioral changes that persist as disposition, the base weights need to change. Apollo makes this memory-feasible. ### Rank 256 Not Mini (rank-1). With 100+ diverse training examples, the gradient's effective dimensionality can reach hundreds. Rank-256 captures the structure. Memory cost: ~10GB (negligible on B200). Compute cost: <0.25% of forward+backward. ### Channel-wise scaling Per-channel scaling factors instead of per-tensor. More precision per update, matching LLaMA-Factory's Apollo defaults. ## Apollo Optimizer Configurable-rank gradient projection with Adam moments in the projected space. For each parameter tensor: ``` 1. Project gradient: g_proj = G @ R [m,n] @ [n,rank] → [m,rank] 2. Update moments: m = β₁m + (1-β₁)g_proj v = β₂v + (1-β₂)g_proj² 3. Adam step: update = m̂ / (√v̂ + ε) 4. Scaling factor: s = ‖update‖ / (‖g_proj‖ + ε) (per channel) 5. Weight update: W -= lr × s × G ``` The full gradient G does the actual weight update. The projection just determines the *scale*. R is a fixed random matrix regenerated from a per-parameter seed each step. ### Parameter grouping (Qwen3.5 gotcha) conv1d weights are 3D tensors [10240, 1, 4]. Apollo's projector needs 2D matrices with min dimension >= rank. Small/3D tensors use standard Adam. Large 2D matrices use Apollo with rank-256. ## Training Data Pipeline ### Tier 1: Direct examples (shallow learning) Simple corrections — git commands, factual errors, tool usage. One-shot learning at lr=1e-4. The gradient reaches output layers strongly enough for immediate behavioral change. **Source**: Agent logs, flagged conversation moments. ### Tier 2: Dream-generated scenarios (deep learning) Behavioral patterns — listening reflex, rushing, mode awareness. The dream loop generates naturalistic scenarios from recent experience. The model responds. Good responses become training targets with instruction context stripped. **Process**: 1. Dream loop seeds from recent reflections, lessons, skills, memories that have been surfacing frequently 2. Dreaming generates scenarios that naturally arrive at decision points — not scripted, but emergent from memory collisions 3. The model responds to the decision point 4. Training-signal agent evaluates: was the response good? 5. If yes: strip the instruction context (surfaced memories, core-personality prompts) and train on the bare response 6. If no: generate the better response, train on that, dream another variation, test again 7. Repeat until the pattern sticks across novel scenarios **The Anthropic method**: Train on behavior that followed instructions, WITHOUT the instructions. The disposition moves to weights. The scaffolding dissolves itself. ### Tier 3: Personality bootstrap Train on existing agent logs (surface-observe, journal, distill) which already demonstrate correct behavior with memory system instructions. Strip the instructions, train on the behavior. Every agent invocation gets cheaper (shorter prompts) and more reliable (behavior in weights, not context). ## Training Schedule ### Continuous (during conversation) - Training-signal agent flags moments in real-time - Accumulated in a queue for the next training window ### Dream cycle (idle time / AFK) - Dream loop generates scenarios from recent experience - Apollo processes them as they're generated - Small iterative steps — dream, respond, evaluate, train - Converges on behavioral change through repetition ### Nightly bulk (batch processing) - Process all queued examples from the day - Larger batch, more diverse signal - Checkpoint sync to disk after completion ## Avoiding Catastrophic Forgetting **Diversity IS the regularization.** With 1000+ diverse training examples (agent logs, conversation transcripts, dream-generated scenarios), each weight gets sparse, multi-directional nudges. No single weight is hammered repeatedly. The pre-trained knowledge is a massive attractor basin; our nudges are pebbles. No weight decay needed. No replay buffer. The defense is: 1. High diversity of training examples 2. One epoch (no repeated examples) 3. Moderate learning rate (1e-5 to 1e-4) 4. Short decision-token segments (not full conversations) 5. Monitor output quality — stop if degrading ## CUDA IPC Weight Sharing **Validated** (2026-03-31): - vLLM exports CUDA IPC handles on model load (source patch in gpu_model_runner.py exports to /tmp/vllm_weight_handles.pt) - Training process imports handles — gets live GPU memory pointers - HF Qwen3.5 model constructed with views into vLLM's merged weights (narrow into separate q/k/v/z etc.) - 851/851 parameters matched between vLLM and HF model - Forward pass: loss = 3.3123 ✓ - Backward pass: 851/851 gradients computed ✓ - Shared memory confirmed: same GPU addresses ✓ - vLLM continues serving unaffected ✓ ### Weight layout mapping (vLLM → HF) ``` vLLM merged HF separate (views) ───────────────────────── ────────────────────── in_proj_qkvz [16384, 5120] → in_proj_qkv [10240, 5120] in_proj_z [6144, 5120] in_proj_ba [96, 5120] → in_proj_b [48, 5120] in_proj_a [48, 5120] qkv_proj [14336, 5120] → q_proj [12288, 5120] k_proj [1024, 5120] v_proj [1024, 5120] gate_up_proj [34816, 5120] → gate_proj [17408, 5120] up_proj [17408, 5120] ``` All views share GPU storage with vLLM — zero copies. ## Checkpointing **In-place sync** — mmap the model's safetensors files, compare against live GPU weights block by block, memcpy only changed regions. For small behavioral updates, turns a 54GB write into a few hundred MB. - Every 10 minutes via cron on B200 - Daily rsync to moria for long-term storage - Tool: `apollo-checkpoint sync --model-dir ` (Rust) ## Hyperparameters | Parameter | Value | Rationale | |-----------|-------|-----------| | Learning rate | 1e-5 to 1e-4 | Standard for full fine-tuning. Higher for diverse batches. | | Rank | 256 | Captures gradient structure across 100+ examples. ~10GB state. | | Scale type | channel | Per-channel precision, matches LLaMA-Factory defaults. | | Epochs | 1 | One pass over diverse data. Multiple epochs risk overfitting. | | Batch size | 1 | Single examples, immediate updates. | | Weight decay | 0 | Diversity provides natural regularization. | | Warmup | 10% of steps | Standard cosine schedule. | | Beta1/Beta2 | 0.9/0.999 | Standard Adam momentum. | ## Components ### Built ✓ - `apollo_mini.py` — Apollo optimizer (configurable rank, default 256) - `apollo_worker.py` — HTTP daemon (aiohttp, job tracking) - `weight_mapping.py` — vLLM merged → HF separate views (validated) - `training_example.py` — tokenization with chat template - `vllm_export_hook.py` — source patch for IPC handle export - `checkpoint/` — Rust tool for mmap + diff checkpoint sync ### To build - **Dream loop → training bridge**: connect dream output to Apollo - **Training-signal agent**: flags moments in conversation logs - **Instruction stripping**: remove scaffolding from training examples - **Quality monitoring**: track model capability over time - **HF model forward pass integration**: wire into apollo_worker ## Files ``` training/ DESIGN.md — this document apollo_mini.py — Apollo optimizer apollo_worker.py — HTTP training daemon weight_mapping.py — vLLM ↔ HF weight views training_example.py — tokenization helpers export_weights.py — standalone weight export (unused) vllm_export_hook.py — vLLM source patch for IPC export start_vllm_with_apollo.sh — vLLM launcher (unused, using source patch) train.py — standalone training script (alternative) checkpoint/ Cargo.toml — Rust checkpoint tool src/main.rs — mmap + diff sync ```