grx-tensor 0.1.0 → 0.2.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- checksums.yaml +4 -4
- data/CHANGELOG.md +17 -0
- data/GUIA_PRINCIPIANTES.md +763 -0
- data/README.es.md +270 -0
- data/README.md +178 -375
- data/grx-tensor.gemspec +4 -2
- data/lib/grx/c_api.rb +41 -37
- data/lib/grx/data.rb +87 -0
- data/lib/grx/loss.rb +55 -25
- data/lib/grx/nn.rb +153 -36
- data/lib/grx/optim.rb +9 -9
- data/lib/grx/serialization.rb +66 -0
- data/lib/grx/storage.rb +15 -24
- data/lib/grx/tensor.rb +198 -65
- data/lib/grx/utils.rb +28 -0
- data/lib/grx/version.rb +1 -1
- data/lib/grx.rb +14 -3
- metadata +8 -3
checksums.yaml
CHANGED
|
@@ -1,7 +1,7 @@
|
|
|
1
1
|
---
|
|
2
2
|
SHA256:
|
|
3
|
-
metadata.gz:
|
|
4
|
-
data.tar.gz:
|
|
3
|
+
metadata.gz: e72486318839399760aded34a10566bfa919ddec52212dc7694b6d4d8c47e843
|
|
4
|
+
data.tar.gz: 7c52178af28a4d75ce1a424539a088270501b0cd5087f0e97c901c228749e827
|
|
5
5
|
SHA512:
|
|
6
|
-
metadata.gz:
|
|
7
|
-
data.tar.gz:
|
|
6
|
+
metadata.gz: 5248ca76ea0f17611cd11c0c883169f17f998fa8270373473559dd1897263192e7c2af3592d7fcca8276798877755edeb058de095cdb8be424388d3b679dcea8
|
|
7
|
+
data.tar.gz: 9f7461cd793bca1347ed006bdde5ff35fdfea968f1706cd03d66d56c0a1d2cbf2cf58036bcceee2096a26747455389f1d0df9f3cd6b4adcf574b5d412c39af5f
|
data/CHANGELOG.md
CHANGED
|
@@ -19,6 +19,23 @@ Versioning follows [Semantic Versioning](https://semver.org/).
|
|
|
19
19
|
|
|
20
20
|
---
|
|
21
21
|
|
|
22
|
+
## [0.2.0] - 2026-08-21
|
|
23
|
+
|
|
24
|
+
### Added
|
|
25
|
+
- **Differentiable Reductions & Operations**: `Tensor#sum` and `Tensor#mean` now create autograd DAG nodes supporting direct `loss.backward`.
|
|
26
|
+
- **Differentiable Loss Functions**: `MSELoss`, `MAELoss`, `BCELoss`, `CrossEntropyLoss`, and `HuberLoss` now return differentiable `Tensor` instances.
|
|
27
|
+
- **Advanced Neural Layers**:
|
|
28
|
+
- `GRX::NN::Embedding` — trainable dense token embeddings with backpropagation.
|
|
29
|
+
- `GRX::NN::LayerNorm` — feature normalization with trainable gamma/beta parameters.
|
|
30
|
+
- **Model Serialization**: High-speed binary `.grx` format with `GRX1` magic header for instant weight saving and loading (`Module#save_weights`, `Module#load_weights`).
|
|
31
|
+
- **Data Pipelines**: `GRX::Data::TensorDataset` and `GRX::Data::DataLoader` for mini-batching, shuffling, and dataset iteration.
|
|
32
|
+
- **Gradient Clipping**: `GRX::Utils.clip_grad_norm!` for stabilizing deep networks and recurrent architectures.
|
|
33
|
+
- **Robust Windows DLL Resolution**: Multi-path search for `grx_core.dll` across multiple fallback directories.
|
|
34
|
+
- **Helper methods**: `GRX.mode` and `GRX.c_loaded?`.
|
|
35
|
+
- **Spanish Documentation**: Complete `README.es.md` and `GUIA_PRINCIPIANTES.md`.
|
|
36
|
+
|
|
37
|
+
---
|
|
38
|
+
|
|
22
39
|
## [0.1.0] - 2026-05-11
|
|
23
40
|
|
|
24
41
|
### Added
|