foxhunt

gitadmin/foxhunt

Fork 0

Commit Graph

Author	SHA1	Message	Date
jgrusewski	22004a7368	refactor(cuda): eliminate candle from ml-core, ml-ppo, and 4 thin crates Hard refactor — no shims, no compat layers. Candle removed from Cargo.toml and all source files in 6 crates: - ml-core: MlDevice enum, checkpoint.rs (safetensors direct), cudarc imports fixed from candle re-export to direct, AdamWConfig lr_decay, cuda_compat gutted. Net -7,341 lines. - ml-ppo: All 16 files rewritten. LSTM→CudaLSTM, VarMap→GpuVarStore, PPOAgent 2306→700 lines, checkpoint→binary format. - ml-ensemble: GPU-resident sigmoid via custom CUDA kernel. - ml-explainability: Integrated gradients via GPU finite-difference kernels. - ml-labeling: Device→MlDevice. - ml-hyperopt: Cargo.toml only. Remaining: ml-dqn (24 files), ml-supervised (4 files), ml crate (104 files). Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>	2026-03-17 22:27:56 +01:00
jgrusewski	4ef9cdb169	refactor(cuda): replace Candle re-export paths with direct cudarc in ml-core CUDA infrastructure in ml-core (cuda_compile, gpu/capabilities, gpu/l2_cache) previously accessed cudarc types through candle_core::cuda_backend::cudarc -- a transitive re-export that coupled low-level CUDA driver calls to Candle. Changes: - Add cudarc 0.17 as direct optional dep (gated behind cuda feature) - Replace all candle_core::cuda_backend::cudarc paths with direct cudarc:: - Generalize OOM detection to accept &dyn Debug (was &CandleError) - Add generic computation_error_to_common_error() alongside legacy alias Candle references: 163 -> 72 (remaining are autograd-dependent: Tensor, Var, GradStore, VarBuilder -- cannot be replaced with cudarc raw ops) Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>	2026-03-17 14:34:58 +01:00
jgrusewski	33cc076eb3	refactor(ml): move compute primitives to ml-core (5b) Move optimizers, gradient_accumulation, gradient_utils, cuda_compat, tensor_ops, and gpu (device config, capabilities, memory profiling) to ml-core. These are shared compute primitives used by all models. Also commit module files for core types (common, config, error, model, traits, types) that were moved from ml to ml-core in task 5a but left staged without being committed. Notable changes: - resolve_batch_size() stays in ml (new batch_size_resolver module) because it depends on memory_optimization::auto_batch_size which has not yet moved to ml-core - FactoredAction legacy bridge converted from inherent impl to extension trait (FactoredActionLegacy) since FactoredAction is now defined in ml-core, not ml - candle-optimisers added to ml-core dependencies (needed by Adam) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>	2026-03-08 15:13:41 +01:00

Author

SHA1

Message

Date

jgrusewski

22004a7368

refactor(cuda): eliminate candle from ml-core, ml-ppo, and 4 thin crates

Hard refactor — no shims, no compat layers. Candle removed from Cargo.toml
and all source files in 6 crates:

- ml-core: MlDevice enum, checkpoint.rs (safetensors direct), cudarc imports
  fixed from candle re-export to direct, AdamWConfig lr_decay, cuda_compat
  gutted. Net -7,341 lines.
- ml-ppo: All 16 files rewritten. LSTM→CudaLSTM, VarMap→GpuVarStore,
  PPOAgent 2306→700 lines, checkpoint→binary format.
- ml-ensemble: GPU-resident sigmoid via custom CUDA kernel.
- ml-explainability: Integrated gradients via GPU finite-difference kernels.
- ml-labeling: Device→MlDevice.
- ml-hyperopt: Cargo.toml only.

Remaining: ml-dqn (24 files), ml-supervised (4 files), ml crate (104 files).

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

2026-03-17 22:27:56 +01:00

jgrusewski

4ef9cdb169

refactor(cuda): replace Candle re-export paths with direct cudarc in ml-core

CUDA infrastructure in ml-core (cuda_compile, gpu/capabilities, gpu/l2_cache)
previously accessed cudarc types through candle_core::cuda_backend::cudarc --
a transitive re-export that coupled low-level CUDA driver calls to Candle.

Changes:
- Add cudarc 0.17 as direct optional dep (gated behind cuda feature)
- Replace all candle_core::cuda_backend::cudarc paths with direct cudarc::
- Generalize OOM detection to accept &dyn Debug (was &CandleError)
- Add generic computation_error_to_common_error() alongside legacy alias

Candle references: 163 -> 72 (remaining are autograd-dependent: Tensor,
Var, GradStore, VarBuilder -- cannot be replaced with cudarc raw ops)

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

2026-03-17 14:34:58 +01:00

jgrusewski

33cc076eb3

refactor(ml): move compute primitives to ml-core (5b)

Move optimizers, gradient_accumulation, gradient_utils, cuda_compat,
tensor_ops, and gpu (device config, capabilities, memory profiling) to
ml-core. These are shared compute primitives used by all models.

Also commit module files for core types (common, config, error, model,
traits, types) that were moved from ml to ml-core in task 5a but left
staged without being committed.

Notable changes:
- resolve_batch_size() stays in ml (new batch_size_resolver module)
  because it depends on memory_optimization::auto_batch_size which
  has not yet moved to ml-core
- FactoredAction legacy bridge converted from inherent impl to
  extension trait (FactoredActionLegacy) since FactoredAction is now
  defined in ml-core, not ml
- candle-optimisers added to ml-core dependencies (needed by Adam)

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

2026-03-08 15:13:41 +01:00

3 Commits