pyrecall

PyPI version CI License: MIT Python 3.11+ PyPI Downloads

Forgetting detection and skill rollback for fine-tuned LLMs.

Fine-tune on new data and your model quietly loses what it already knew — coding ability, reasoning, safety guardrails. pyrecall catches it before it ships.

Install

pip install pyrecall   # Python 3.11–3.14 · CUDA, MPS, and CPU

Quickstart

from pyrecall import Model

model = Model("meta-llama/Llama-3.2-1B")
model.snapshot("before")
model.learn("data.jsonl", epochs=3)

if not model.check().is_healthy:
    model.rollback(to="before")

pyrecall init --model meta-llama/Llama-3.2-1B
pyrecall snapshot before_v1
pyrecall learn train.jsonl --snapshot-after after_v1
pyrecall check --before before_v1 --after after_v1
# exit 0 → ship   exit 2 → pyrecall rollback before_v1

How it works

Benchmarks 180 prompts across 9 skill categories (reasoning, coding, safety, math, multilingual, and more) using log-likelihood scoring. After training, check diffs the scores and flags any category that drops past your threshold (default 10%). Only the LoRA adapter is stored per snapshot — a few hundred MB, not the full model.

Any causal LM on HuggingFace Hub is supported — Llama, Mistral, Phi, Gemma, Qwen, Falcon, GPT-2/Neo/J, and more. LoRA targets are auto-detected.

Docs

Full CLI reference, Python API, experiment tracker integrations (W&B, MLflow, Neptune), custom benchmarks, per-category thresholds, and more:

pyrecall.github.io/Pyrecall

Contributing

git clone https://github.com/Arths17/Pyrecall
pip install -e ".[dev]"
pytest

Open an issue before large changes. MIT — LICENSE.

Project links

Data verified by PyPI on Jun 17, 2026
Data provided by the project maintainers, verified at the time the release was uploaded to PyPI.

Key dates

PyPI data
Data sourced directly from PyPI's database.
  • Released:
Newer release available (0.12.1)

2 maintainers

PyPI data
Data sourced directly from PyPI's database.
Avatar for AtharvRanjan from gravatar.com AtharvRanjan Avatar for Sid907 from gravatar.com Sid907

Credits

Author: pyrecall contributors

GitHub Statistics

Data verified by PyPI on Jun 17, 2026
The GitHub source repository was provided by the project maintainers and verified by PyPI at the time of upload. Stars, forks, and open issues/PRs are derived from that repository and have not been independently verified.

License

MIT License (MIT License)

Requires

Python >=3.11

Provides Extra

dev mlflow neptune privacy quantization serve trackers wandb

Tags

catastrophic-forgetting continual-learning fine-tuning llm lora peft transformers

Classifiers

Report project as malware

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

pyrecall-0.11.10.tar.gz (152.2 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

pyrecall-0.11.10-py3-none-any.whl (87.6 kB view details)

Uploaded Python 3

File details

Details for the file pyrecall-0.11.10.tar.gz.

File metadata

  • Download URL: pyrecall-0.11.10.tar.gz
  • Upload date:
  • Size: 152.2 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for pyrecall-0.11.10.tar.gz
Algorithm Hash digest
SHA256 4a15bd815014085a6183b8cd7d8d4329bd3149b0e7cff55d3c05106ff0b5e0b0
MD5 8d7cab581e2db5a26c72a91da4cf9e17
BLAKE2b-256 c6e2c39e122ba8154d1fc0215abfbd58211755ec6a83411e61011f2d5db8c62f

See more details on using hashes here.

Provenance

The following attestation bundles were made for pyrecall-0.11.10.tar.gz:

Publisher: publish.yml on Pyrecall/Pyrecall

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file pyrecall-0.11.10-py3-none-any.whl.

File metadata

  • Download URL: pyrecall-0.11.10-py3-none-any.whl
  • Upload date:
  • Size: 87.6 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for pyrecall-0.11.10-py3-none-any.whl
Algorithm Hash digest
SHA256 c3ed41cf32ebb9fea0c4e638eddbf07c0e225e26eb93df48a45360138fdf4dbc
MD5 8da8d96726a9aa06e407b2a2ed33c150
BLAKE2b-256 a618ea3bd9da6781134e97aff5b3dac9198391f01700d2f9cfda4c0694d34935

See more details on using hashes here.

Provenance

The following attestation bundles were made for pyrecall-0.11.10-py3-none-any.whl:

Publisher: publish.yml on Pyrecall/Pyrecall

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Release history Release notifications | RSS feed

0.12.1

2 files

0.12.0

2 files

0.11.11

2 files

This release

0.11.10 This release

2 files

0.11.9

2 files

0.11.8

2 files

0.11.7

2 files

0.11.6

2 files

0.11.5

2 files

0.11.4

2 files

0.11.3

2 files

0.11.2

2 files

0.11.1

2 files

0.11.0

2 files

0.10.1

2 files

0.10.0

2 files

0.9.0

2 files

0.8.0

2 files

0.7.0

2 files

0.6.0

2 files

0.5.0

2 files

0.4.0

2 files

0.3.0

2 files

0.2.1

2 files

0.2.0

2 files

0.1.6

2 files

0.1.5

2 files

0.1.2

2 files

0.1.1

2 files

0.1.0

2 files

0.0.1

2 files

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page