jgrusewski e12cdfa005 feat(ml): diffusion trainer GPU-accumulated loss + batched grad norm ...

Replace per-param gradient norm extraction with batched Tensor::stack
pattern (N GPU syncs → 1). Remove per-batch loss scalar extraction
from backward(), defer to epoch-level accumulation. Apply GPU-accumulated
validation loss (stack + mean_all instead of per-batch to_scalar).

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

2026-03-02 17:30:19 +01:00

..

backtesting

docs: create/update README.md for all 17 crates

2026-03-01 22:47:39 +01:00

common

refactor(common): make init_observability sync with optional OTLP endpoint

2026-03-01 23:03:13 +01:00

config

docs: create/update README.md for all 17 crates

2026-03-01 22:47:39 +01:00

ctrader-openapi

docs: create/update README.md for all 17 crates

2026-03-01 22:47:39 +01:00

data

docs: create/update README.md for all 17 crates

2026-03-01 22:47:39 +01:00

database

docs: create/update README.md for all 17 crates

2026-03-01 22:47:39 +01:00

market-data

docs: create/update README.md for all 17 crates

2026-03-01 22:47:39 +01:00

ml

feat(ml): diffusion trainer GPU-accumulated loss + batched grad norm

2026-03-02 17:30:19 +01:00

ml-data

docs: create/update README.md for all 17 crates

2026-03-01 22:47:39 +01:00

model_loader

docs: create/update README.md for all 17 crates

2026-03-01 22:47:39 +01:00

risk

docs: create/update README.md for all 17 crates

2026-03-01 22:47:39 +01:00

risk-data

docs: create/update README.md for all 17 crates

2026-03-01 22:47:39 +01:00

storage

docs: create/update README.md for all 17 crates

2026-03-01 22:47:39 +01:00

trading_engine

docs: create/update README.md for all 17 crates

2026-03-01 22:47:39 +01:00

trading-data

docs: create/update README.md for all 17 crates

2026-03-01 22:47:39 +01:00

training_uploader

docs: add lib.rs doc comment to training_uploader

2026-03-01 22:53:11 +01:00

web-gateway

refactor(common): make init_observability sync with optional OTLP endpoint

2026-03-01 23:03:13 +01:00