imcaspar/gpt2-ml

Health v2 · maintenance only

Adoption not part of the score

1701 stars · 325 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

GPT2-ML is a TensorFlow-based GPT-2 training and inference project supporting multiple languages, with a focus on Chinese. It provides 1.5B-parameter pretrained Chinese models trained on ~30G of corpus using Cloud TPUs, plus a Colab demo notebook.

Use cases

• generate Chinese text with a pretrained GPT-2 model
• fine-tune GPT-2 on Chinese corpora using TPUs
• try a 1.5B Chinese language model in Google Colab
• train a multilingual GPT-2 model with a BERT tokenizer
• download pretrained Chinese GPT-2 checkpoints for NLP research

When to choose

• you need a large pretrained Chinese GPT-2 model
• you want TPU-based GPT-2 training scripts in TensorFlow
• you want a quick Colab demo of Chinese text generation

When to avoid

• you need PyTorch-based tooling or modern LLM inference stacks
• you need actively maintained code with recent updates
• you need multilingual coverage beyond Chinese in practice

Facets

library · maturity maintenance

machine-learning deep-learning llm-training nlp large-language-models machine-learning python cloud gpt-2 pretrained-models chinese text-generation tensorflow tpu bert-tokenizer colab natural-language-processing gpu

Member repositories

For agents

markdown · JSON · MCP: product_card(name="imcaspar/gpt2-ml")