shiyu-coder/Kronos
43.7
Weak · 15 November 2026
8.3k
lines of production code
Python
with JavaScript
5
measurements over time
What this system is
This system is a codebase dedicated to machine learning fine-tuning workflows, specifically focusing on data processing and model training pipelines. Recent development has centered on establishing a custom CSV dataset fine-tuning process and refining the tokenizer training logic to handle batch dimensions correctly. The project is currently in a phase of stabilizing these core training components through targeted behavioral adjustments and expanded test coverage.
How it got here
October 2025 — CSV fine-tuning and test coverage
3 changes.
The period focused on establishing a custom CSV dataset fine-tuning workflow, supported by the addition of example kline data for testing. This development work was accompanied by the addition and updating of tests to ensure coverage for the new functionality.
April 2026 — Finetune tokenizer and behavioral changes
2 changes.
This period focused on refining the finetune pipeline, specifically addressing batch dimension handling within the tokenizer training process. The majority of activity involved broader behavioral changes across the repository, with significant effort directed toward modifying finetune logic and resolving associated issues.
Written by cca.canine.dev from the codebase's own history, inside the signed delivery this page is composed from.
How this codebase got here
Score
- CAI 44 → 44 (-0.2)
- Rubric changed (rubric-2026.11.1 → rubric-2026.11.4) — scores are not directly comparable.
Lenses
- Code Health 64 → 56 (-7.5)
- Architecture 96 → 96 (+0.0)
- Maturity 45 → 44 (-0.8)
- Readiness 36 → 36 (+0.0)
- Security 48 → 47 (-0.8)
- Accessibility 53 → 60 (+7.5)
Resolved (16)
- Duplicated block (10 lines × 2) (finetune/train_predictor.py)
- Duplicated block (12 lines × 2) (examples/prediction_new.py)
- Duplicated block (13 lines × 2) (examples/get_date_new.py)
- Duplicated block (13–14 lines × 2) (examples/get_akshare_date_2024-2025_x.py)
- Duplicated block (13–15 lines × 2) (webui/app.py)
- Duplicated block (20–23 lines × 2) (examples/run_backtest_kronos.py)
- Duplicated block (26 lines × 2) (finetune_csv/finetune_base_model.py)
- Duplicated block (50 lines × 2) (examples/prediction_new.py)
- Duplicated block (6 lines × 2) (examples/get_akshare_date_2024-2025_x.py)
- Duplicated block (6 lines × 2) (finetune_csv/train_sequential.py)
- Duplicated block (6 lines × 3) (examples/get_date_new.py)
- Duplicated block (7 lines × 2) (examples/prediction_new.py)
- Near-duplicate member pair (101 shared lines) (examples/get_akshare_date_2024-2025_x.py)
- Near-duplicate member pair (23 shared lines) (examples/prediction_batch_example.py)
- Near-duplicate member pair (51 shared lines) (examples/prediction_new.py)
- Off-boarding risk: anonymized user #1
New (20)
- Edited copy of a member (113 corresponding lines) (examples/get_akshare_date_2024-2025_x.py)
- Edited copy of a member (119 corresponding lines) (finetune/train_predictor.py)
- Edited copy of a member (121 corresponding lines) (examples/get_akshare_date_2024-2025_x.py)
- Edited copy of a member (125 corresponding lines) (finetune_csv/finetune_base_model.py)
- Edited copy of a member (150 corresponding lines) (examples/prediction_new.py)
- Edited copy of a member (16 corresponding lines) (examples/prediction_akshare_2024-2025.py)
- Edited copy of a member (23 corresponding lines) (model/module.py)
- Edited copy of a member (25 corresponding lines) (finetune/dataset.py)
- Edited copy of a member (28 corresponding lines) (finetune/train_predictor.py)
- Edited copy of a member (30 corresponding lines) (examples/prediction_batch_example.py)
- Edited copy of a member (39 corresponding lines) (examples/prediction_new.py)
- Edited copy of a member (40 corresponding lines) (examples/get_akshare_date_2024-2025_x.py)
- Edited copy of a member (42 corresponding lines) (examples/get_akshare_date_2024-2025_x.py)
- Edited copy of a member (49 corresponding lines) (examples/prediction_new.py)
- Edited copy of a member (50 corresponding lines) (examples/get_akshare_date_2024-2025_x.py)
- Edited copy of a member (53 corresponding lines) (finetune/train_predictor.py)
- Edited copy of a member (55 corresponding lines) (examples/prediction_new.py)
- Edited copy of a member (7 corresponding lines) (finetune/utils/training_utils.py)
- Edited copy of a member (8 corresponding lines) (finetune/utils/training_utils.py)
- Edited copy of a member (8 corresponding lines) (model/module.py)
Written by cca.canine.dev from the codebase's own history, inside the signed delivery this page is composed from.
Survey your own repository
shiyu-coder/Kronos was measured the same way every project in this corpus was: the same rubric, at a pinned commit, with the result published in full. Point a surveyor at a repository you know and see whether you agree with it.
About this page
- The score is its most recent published measurement, taken on 15 November 2026 at a pinned commit. It is not a live figure and does not change until the project is measured again.
- Measured at commit 67b630e67f6a18c9e9be918d9b4337c960db1e9a — the exact code this score is about.
- Scored under rubric-2026.11.4 — the same rubric and the same method as every other entry in this index.
- Measured by Canine Code Analyzer (cca.canine.dev) using cca measure-2026.11.73+preprod-52de90b74216.