Model.Close() now walks the full model tree (Gemma3/Qwen3) and
explicitly frees all weight arrays. Handles tied output weights,
nil safety, idempotent double-close. Helpers: freeLinear,
freeEmbedding, freeRMSNorm, freeCaches, closeGemma, closeQwen3.
Array.Iter() returns iter.Seq[float32] for range-over-func iteration.
Handles non-contiguous arrays and supports early break.
192 tests passing (12 new: 8 close, 4 iter).
Co-Authored-By: Virgil <virgil@lethean.io>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Version bump validation, batch evaluation patterns, GPU profiling,
quantised matmul variants, and async/streaming patterns — all
activate when the Go side needs them.
Co-Authored-By: Virgil <virgil@lethean.io>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Replace checkError() log+swallow with lastError() that returns real MLX
error messages. Add Eval/EvalAsync as error-returning variants of
Materialize. Generate loop now propagates GPU errors via model.Err().
LoadAllSafetensors returns (map, error). Model loaders check lastError()
after safetensors load. 180 tests passing.
Co-Authored-By: Virgil <virgil@lethean.io>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
CLion Claude session 1 completed all orientation and priority tasks:
- Full API surface map (180 ops, Go binds ~40)
- mlx_contiguous found (fixes Floats() bug)
- mlx_cumsum found (unblocks TopP sampling)
- Memory management fully documented (refcounted, safe to free)
- Error model documented (free-form strings only)
- Device info API documented (mlx_metal_device_info)
Co-Authored-By: Virgil <virgil@lethean.io>
New bindings from mlx-c memory.h and metal.h:
- GetCacheMemory() — current allocator cache size
- ResetPeakMemory() — reset high-water mark
- SetWiredLimit() — control wired memory limit
- GetDeviceInfo() — GPU architecture, max buffer, memory size
All exposed at root package level via register_metal.go delegates.
Co-Authored-By: Virgil <virgil@lethean.io>
New ops: CumSum, Sort, Argsort, Greater, MaxAxis — all bound to mlx-c.
TopP (nucleus) sampling now fully implemented: sorts probabilities
descending, computes cumulative sum, masks tokens beyond the threshold,
and scatters the mask back to original positions via argsort.
MinP sampling now fully implemented: computes softmax, finds max
probability, masks tokens below min_p * max_prob.
Both were previously stubs that passed through logits unchanged.
10 new tests (CumSum variants, Sort, Argsort, Greater, MaxAxis,
TopP, MinP). 176 total tests passing.
Co-Authored-By: Virgil <virgil@lethean.io>
Bind mlx_contiguous and _mlx_array_is_row_contiguous from mlx-c.
Floats(), DataInt32(), and Ints() now automatically handle non-contiguous
arrays (from Transpose, BroadcastTo, SliceAxis, etc.) by checking
IsRowContiguous() and making a contiguous copy when needed.
Previously these methods returned silently wrong data for view arrays.
The old workaround of Reshape(arr, totalSize) is no longer needed.
7 new tests for contiguous handling (transpose, broadcast, slice views).
Co-Authored-By: Virgil <virgil@lethean.io>
Replace local TextModel, Backend, Token, Message, and option types with
forge.lthn.ai/core/go-inference. go-mlx is now a pure backend that
registers "metal" into the shared inference registry via init().
Deleted: textmodel.go, options.go, backend.go
Updated: register_metal.go (implements inference.Backend with Available()),
mlx_test.go (uses inference.* types, 4 new tests), go.mod,
internal/metal/generate.go (added RepeatPenalty)
159 tests passing (148 internal/metal + 11 root).
Co-Authored-By: Virgil <virgil@lethean.io>
Tests: MetalAvailable, DefaultBackend, GetBackend, LoadModel error
paths, options, defaults. Model-dependent test skips when model
not available on disk.
Co-Authored-By: Virgil <virgil@lethean.io>
Move foundation CGO files from root package to internal/metal/ package.
Changes package declaration from `package mlx` to `package metal`.
Updates CGO SRCDIR paths to account for new location (two levels deeper).
Extracts go:generate directives into root generate.go.
Co-Authored-By: Virgil <virgil@lethean.io>
Virgil review items integrated:
- context.Context on Generate/Chat (required for HTTP cancellation)
- Err() error on TextModel (distinguish EOS from OOM)
- Chat() on TextModel (model owns its chat template)
- Memory control functions exposed at root package level
- Functional options convention confirmed
- pkg/process confirmed — no changes needed for mlxlm
Co-Authored-By: Virgil <virgil@lethean.io>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Approved design for restructuring go-mlx:
- Root package becomes clean interface (TextModel, LoadModel, Token)
- All CGO code moves to internal/metal/
- Deterministic memory management (Close + per-step cleanup)
- Error propagation instead of silent logging
- mlxlm/ backend placeholder for Python subprocess support
Includes API breaking change communication in FINDINGS.md and
memory management research tasks in cpp/TODO.md.
See: docs/plans/2026-02-19-backend-abstraction-design.md
Co-Authored-By: Virgil <virgil@lethean.io>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
cpp/ now has its own CMakeLists.txt that independently fetches mlx-c
v0.4.1, allowing a CLion Claude to build and explore the C API surface
without conflicting with the Go-side CMake build. Updated CLAUDE.md to
reflect cpp/ as the project root with standalone build instructions.
Added priority tasks to cpp/TODO.md: mlx_contiguous search, eval
semantics verification, mlx_cumsum check for TopP sampling.
Co-Authored-By: Virgil <virgil@lethean.io>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Split mlx/ directory from forge.lthn.ai/core/go-ai into its own module.
Rewrites import paths, adds CLAUDE.md/TODO.md/FINDINGS.md for dedicated
Claude sessions. Zero external Go deps — pure CGO + mlx-c v0.4.1.
Co-Authored-By: Virgil <virgil@lethean.io>
