3bef72f50b
feat(model): add TimeAwareGPT2TemporalConv using TemporalConvEncoder; wire into train.py and utils.load_model; add model_name to configs; translate CN comments and add math import
2025-10-22 17:34:06 +08:00
a81da36657
R: order ICD-10 chapters by chapter number (Roman numeral prefix) in boxplots; handle 'Death' and unknowns at end
2025-10-22 16:08:00 +08:00
b954b4b3e7
R: add orientation arg to boxplot script; toggle coord_flip for rotated (horizontal) vs vertical layouts and adjust axis labels
2025-10-22 15:58:15 +08:00
f8e0104d6b
R: add plot_auc_boxplots_by_chapter.R to compare AUC distributions by ICD-10 chapter for 1-year and no-gap datasets; outputs individual and grid plots
2025-10-22 15:52:04 +08:00
262a7db0da
R: add plot_model_comparison_no_gap.R mirroring 1-year script; generate per-plot and 2x2 cowplot grid from model_comparison_auc_no_gap.csv
2025-10-22 15:43:16 +08:00
9917b3ab63
R: extend 1-year AUC plot script with additional comparisons (256_l, 120_l vs Delphi; 120 vs 120_l; 256 vs 256_l) and generic xy helper
2025-10-22 15:32:02 +08:00
8316326d7e
Add R script to plot 1-year AUC comparisons with ggplot2 (256 vs Delphi and 120 vs Delphi)
2025-10-22 15:25:58 +08:00
6dd5eb95c7
evaluate_models.ipynb: update load_model usage to new signature (infer checkpoint from config)
2025-10-22 13:27:28 +08:00
5b0642eb6e
Add train_ddp.py: DistributedDataParallel multi-GPU training with distributed samplers, rank-0 checkpointing, and aggregated metrics
2025-10-22 11:54:48 +08:00
93cf2018d2
evaluate_auc: adapt to new utils.load_model signature (infer checkpoint from config)
2025-10-22 11:40:46 +08:00
6801e5bdbb
utils.load_model: take only config; infer checkpoint name from config (with legacy fallback) and vocab from checkpoint
2025-10-22 11:39:10 +08:00
92a5bd4a83
evaluate_auc: use new utils.load_model (infer vocab, model variants) and dynamic device; remove unused imports
2025-10-22 11:35:52 +08:00
dfdf64da9a
Rewrite load_model to match train.py: support model variants and infer vocab_size from checkpoint; load state dict robustly
2025-10-22 11:30:59 +08:00
bd88daa8c2
update models and training scripts
2025-10-22 08:36:55 +08:00
e348086e52
update
2025-10-21 10:30:18 +08:00
a8aa5a2bd6
update
2025-10-21 09:20:43 +08:00
ddb7dbfc67
update
2025-10-20 16:22:15 +08:00
88cccdad2e
feat: Optimize AUC evaluation with parallel processing
2025-10-20 16:16:50 +08:00
8f44018bae
update evaluate
2025-10-20 13:47:50 +08:00
1c9e2a2fb3
feat: print model config and add evaluation notebook
2025-10-20 10:14:50 +08:00
6b782b86e1
feat: Add model checkpoints and configurations
2025-10-20 09:38:24 +08:00
9a9de170d1
delete
2025-10-18 22:35:42 +08:00
7e57e5d3b1
refactor: Update survival loss calculation in CombinedLoss
2025-10-18 15:21:10 +08:00
14865ac5b6
Refactor: Remove Jupyter Notebook cell markers
2025-10-18 13:32:26 +08:00
dbc3000192
add evaluation scripts.
2025-10-18 13:26:56 +08:00
082c719975
feat(models): Refactor generate function in TimeAwareGPT2 with competing risks sampling
2025-10-18 12:42:14 +08:00
a631ac6d59
feat: Add load_model function and update training script
...
Added a `load_model` function to `utils.py` to allow loading of trained models from configuration and state dictionary files.
The `train_iter.py` script was also modified, likely to incorporate or test this new functionality.
2025-10-18 11:07:59 +08:00
f7356b183c
feat: Add command-line arguments to train.py
2025-10-18 10:23:12 +08:00
3390bc025e
feat: Add iteration-based training scripts (single and multi-GPU)
2025-10-18 10:05:37 +08:00
a832a45c62
config: Tune hyperparameters for multi-GPU training
...
Increase model size (n_embd, n_layer, n_head) for the multi-GPU configuration.
Explicitly set AdamW betas to (0.9, 0.99).
2025-10-17 15:37:42 +08:00
d760c45baf
feat: Add multi-GPU training and improve config/ignore
...
Add train_multigpu.py for distributed data parallel training.
Update train.py to save the training configuration to a JSON file.
Generalize .gitignore to exclude all *.pt checkpoint files.
Delete obsolete train_dpp.py file.
2025-10-17 14:09:34 +08:00
053f86f4da
config: Add weight decay to training configuration
...
Adds a weight_decay parameter to the TrainConfig and applies it to the AdamW optimizer.
2025-10-17 13:47:37 +08:00
d4d25ac9c7
feat: Add covariate-aware model and piecewise encoder
...
Introduce PiecewiseLinearEncoder for continuous variable encoding.
Add CovariateAwareGPT2 to extend TimeAwareGPT2 with static and time-varying covariate processing.
The model combines piecewise linear and sinusoidal encodings for covariates and integrates them via concatenation before a final MLP head.
Reorganize models.py for better logical structure.
2025-10-17 12:04:50 +08:00
fe0304a96a
feat: Save model with params in name and log losses
2025-10-17 10:44:17 +08:00
7e8d8d307b
chore: Ignore small data files
2025-10-17 10:34:24 +08:00
fc0aef4e71
chore: Add .gitignore
2025-10-17 10:32:42 +08:00
02d84a7eca
refactor: Use AdamW optimizer and increase early stopping patience
2025-10-17 10:31:12 +08:00
cb7575a229
feat: Update model and training parameters
...
In `models.py`:
- Change temporal attention mask to be strictly causal (`<` instead of `<=`).
- Add self-attention for the first token in a sequence to prevent NaNs.
In `train.py`:
- Update hyperparameters:
- `block_length`: 24 -> 48
- `n_embd`: 256 -> 120
- `n_layer`: 8 -> 12
- `n_head`: 8 -> 12
2025-10-16 18:50:15 +08:00
e2495f43b0
revert 6e0713048a
...
revert update attn mask
2025-10-16 18:37:55 +08:00
6e0713048a
update attn mask
2025-10-16 18:29:48 +08:00
eec406d79f
update ignored events.
2025-10-16 17:10:01 +08:00
e3e533c9ec
update
2025-10-16 16:58:30 +08:00
b5172392cb
update dpp
2025-10-16 16:46:33 +08:00
6b0b86d9d0
add Multi_GPU support.
2025-10-16 16:28:52 +08:00
c7296381b8
Revert "feat: adapt train.py to multi-GPU environment"
...
This reverts commit b7aad7a774
2025-10-16 16:23:38 +08:00
2b20299e36
Revert "fix: average loss for multi-GPU training"
...
This reverts commit 85502561ee
2025-10-16 16:23:35 +08:00
85502561ee
fix: average loss for multi-GPU training
2025-10-16 16:21:51 +08:00
b7aad7a774
feat: adapt train.py to multi-GPU environment
2025-10-16 16:16:15 +08:00
4181ead03a
Refactor: Improve attention mechanism and early stopping
...
- Refactor the self-attention mechanism in `models.py` to use `nn.MultiheadAttention` for better performance and clarity.
- Disable early stopping check during warmup epochs in `train.py` to improve training stability.
2025-10-16 15:57:27 +08:00
8a757a8b1d
feat: Add training and validation data via Git LFS
2025-10-16 14:24:56 +08:00
