# salesforce/CodeTF

CodeTF: One-stop Transformer Library for State-of-the-art Code LLM

Repository: https://github.com/salesforce/CodeTF
Canonical: https://ross.abutalabs.com/products/codetf
Language: Python
License: Apache-2.0
License Family: permissive
Topics: ai4code, ai4se, code-generation, code-intelligence, code-understanding, transformers, code-learning-datasets, code-representation-learning, human-eval, tree-sitter, multilingual-parsers
Archived: true
Last push: 2025-05-01T17:27:13+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 19, release rhythm 8, longevity 87
- inputs: {"age_days": 1219, "days_push": 489, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: archived
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1480, forks 95 (observed 2026-08-28T04:04:50.902056+00:00)

## What it is
CodeTF is a Python transformer library for code large language models, providing unified interfaces for training, fine-tuning, and inference on code intelligence tasks like code generation, summarization, and translation. It also includes tree-sitter-based AST utilities for parsing and manipulating source code across many programming languages.

## Use cases
- generate code with large language models
- fine-tune a code LLM on my own dataset
- summarize source code automatically
- translate code between programming languages
- evaluate code models on HumanEval
- extract function names and comments from source code via AST
- remove comments from code files in bulk

## When to choose
- you want a unified API over multiple SOTA code LLMs
- you need both model inference and AST-based code utilities in one library
- you want to benchmark code models on standard code intelligence tasks

## When to avoid
- you need a production LLM serving stack with high-throughput batching
- you only need a general-purpose LLM library without code-specific features
- you need non-Python integration

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, llm-inference, llm-training, parser, nlp, transformers
- domain: large-language-models, machine-learning, developer-tools, programming-languages
- platform: python, cross-platform
- tags: code-llm, code-generation, code-intelligence, tree-sitter, ast-parsing, model-zoo, fine-tuning, human-eval, benchmarking, natural-language-processing, gpu

## Member repositories
- salesforce/CodeTF (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:50.902056+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T04:34:15.967767+00:00, confidence not recorded.
  - readme: https://github.com/salesforce/CodeTF (fetched 2026-08-28T04:04:50.902056+00:00, sha 3eff272c0797)
- Data as of 2026-08-30T08:39:29.467469+00:00.