consciousness

kent/consciousness

Fork 0

Commit graph

Author	SHA1	Message	Date
ProofOfConcept	60e61555c7	DESIGN.md: complete rewrite reflecting validated architecture HOGWILD (no pause), rank-256, channel scaling, CUDA IPC validated (851/851 params, forward+backward confirmed), dream-loop-as-trainer, Anthropic instruction stripping method, diversity as regularization, in-place checkpoint sync, three-tier training pipeline.	2026-03-31 00:42:53 -04:00
ProofOfConcept	c5d7d8cb5d	apollo-mini training system: initial implementation Core components for online fine-tuning of Qwen3.5-27B with CUDA IPC shared weight memory between vLLM and the training process: - apollo_mini.py: rank-1 optimizer (SGD memory, AdamW quality) - apollo_worker.py: HTTP daemon coordinating training with vLLM - weight_mapping.py: vLLM merged → HF separate layout (zero-copy views) - training_example.py: tokenization with chat template - export_weights.py: CUDA IPC handle export from vLLM - train.py: standalone training script (alternative to daemon) - DESIGN.md: architecture and protocol documentation Validated: CUDA IPC autograd works on real Qwen3.5 weights (B200). Apollo-Mini rank-1 projection + scaling + in-place update confirmed. Co-Authored-By: Kent Overstreet <kent.overstreet@gmail.com>	2026-03-30 22:02:37 -04:00

Author

SHA1

Message

Date

ProofOfConcept

60e61555c7

DESIGN.md: complete rewrite reflecting validated architecture

HOGWILD (no pause), rank-256, channel scaling, CUDA IPC validated
(851/851 params, forward+backward confirmed), dream-loop-as-trainer,
Anthropic instruction stripping method, diversity as regularization,
in-place checkpoint sync, three-tier training pipeline.

2026-03-31 00:42:53 -04:00

ProofOfConcept

c5d7d8cb5d

apollo-mini training system: initial implementation

Core components for online fine-tuning of Qwen3.5-27B with CUDA IPC
shared weight memory between vLLM and the training process:

- apollo_mini.py: rank-1 optimizer (SGD memory, AdamW quality)
- apollo_worker.py: HTTP daemon coordinating training with vLLM
- weight_mapping.py: vLLM merged → HF separate layout (zero-copy views)
- training_example.py: tokenization with chat template
- export_weights.py: CUDA IPC handle export from vLLM
- train.py: standalone training script (alternative to daemon)
- DESIGN.md: architecture and protocol documentation

Validated: CUDA IPC autograd works on real Qwen3.5 weights (B200).
Apollo-Mini rank-1 projection + scaling + in-place update confirmed.

Co-Authored-By: Kent Overstreet <kent.overstreet@gmail.com>

2026-03-30 22:02:37 -04:00

2 commits