Repository navigation
Sequence Diagrams
Sudhanshu edited this page Nov 11, 2026
·
1 revision
sequenceDiagram
autonumber
participant User as Python/Jupyter user
participant Data as Dataset + split
participant Tok as AutoTokenizer
participant Model as DistilBERT classifier
participant Tr as HF Trainer
participant Disk as Local artifacts
User->>Data: Create 30 labeled examples; stratified 80/20 split
Data->>Tok: map(tokenize_function)
Tok-->>Data: token IDs and attention masks
User->>Model: from_pretrained(distilbert-base-uncased, 4 labels)
User->>Tr: create_trainer(model, args, datasets, collator, metrics)
loop 4 epochs; batches of 8
Tr->>Model: forward(batch)
Model-->>Tr: logits / loss
Tr->>Model: backward + optimizer step
opt epoch evaluation
Tr->>Model: validation forward
Model-->>Tr: predictions
end
end
Tr->>Disk: save_model(best_model)
User->>Tok: save_pretrained(best_model)
User->>Tr: evaluate(), predict()
Tr-->>User: accuracy, weighted F1, report/confusion matrix
flowchart LR
A[Training run] --> B[support_classifier checkpoints]
B --> C[support_classifier/best_model]
C --> D[Model weights + config]
C --> E[Tokenizer files]
A --> F[support_classifier_frozen runs]
G[Proposed Phi-2 SFT] -.-> H[sft_output / sft_model]
H -.-> I[dpo_output / dpo_model]
J[.gitignore] -.->|excludes large weights, cache, venv| B
For the alternate frozen-backbone run, a separate Trainer is constructed after the full model's predictions are printed. The intended Phi-2 SFT/DPO sequence is available on Phi2-SFT-DPO-Design.
Transformer Training Wiki · Implementation status: DistilBERT source available.This project is licensed under the Apache License, Version 2.0.© 2026 Sudhanshu Shukla · Licensed under the Apache License, Version 2.0
Overview
Architecture
Implemented classifier
- DistilBERT-Workflow
- Dataset-and-Labels
- Training-Configuration
- Evaluation-and-Inference
- Frozen-Backbone-Comparison
- Artifacts-and-Checkpoints
Experimental alignment draft
Operations