zj360202 / models

A collection of large language models for natural language generation.

Models

LLMs for NLG (in German)

This repository provides an comprehensive overview of available large language models (LLM) for natural language generation (NLG). Of particular interest to us are models with German language capabilities.

Name	Size	Model Card	License	Implementation	Paper
GPT-J	6B	EleutherAI/gpt-j-6B	MIT, Weights: Apache 2.0	https://github.com/kingoflolz/mesh-transformer-jax/	-
BLOOMZ	560M 1.1B 1.7B 3B 7.1B 176B	bigscience/bloomz-560m bigscience/bloomz-1b1 bigscience/bloomz-1b7 bigscience/bloomz-3b bigscience/bloomz-7b1 bigsience/bloomz	RAIL	https://github.com/bigscience-workshop/xmtf	Muennighoff, Niklas, et al. Crosslingual generalization through multitask finetuning. (2022). DOI: https://doi.org/10.48550/arXiv.2211.01786
BLOOM german	350M 1.5B 6.4B	malteos/bloom-350m-german malteos/bloom-1b5-clp-german malteos/bloom-6b4-clp-german	RAIL	https://github.com/malteos/clp-transfer	Ostendorff, Malte; Rehm, Georg. Efficient Language Model Training through Cross-Lingual and Progressive Transfer Learning. (2023). DOI: https://doi.org/10.48550/arXiv.2301.09626
OPT	125M 350M 1.3B 2.7B 6.7B 13B 30B 66B	facebook/opt-125m facebook/opt-350m facebook/opt-1.3b facebook/opt-2.7b facebook/opt-6.7b facebook/opt-13b facebook/opt-30b facebook/opt-66b	OPT-LICENSE	https://github.com/facebookresearch/metaseq	Zhang, Susan, et al. Opt: Open pre-trained transformer language models. (2022). DOI: https://doi.org/10.48550/arXiv.2205.01068
FLAN-T5	80M 250M 780M 3B 11B	google/flan-t5-small google/flan-t5-base google/flan-t5-large google/flan-t5-xl google/flan-t5-xxl	Apache 2.0	https://github.com/google-research/t5x	Chung, Hyung Won, et al. Scaling instruction-finetuned language models. (2022). DOI: https://doi.org/10.48550/arXiv.2210.11416
MT0	300M 580M 1.2B 3.7B 13B	bigscience/mt0-small bigscience/mt0-base bigscience/mt0-large bigscience/mt0-xl bigscience/mt0-xxl	Apache 2.0	https://github.com/bigscience-workshop/xmtf	Muennighoff, Niklas, et al. Crosslingual generalization through multitask finetuning. (2022). DOI: https://doi.org/10.48550/arXiv.2211.01786
GPT2	117M 117M 1.5B ? ? ?	benjamin/gpt2-wechsel-german malteos/gpt2-wechsel-german-ds-meg malteos/gpt2-xl-wechsel-german benjamin/gerpt2 benjamin/gerpt2-large dbmdz/german-gpt2	MIT	https://github.com/CPJKU/wechsel https://github.com/bminixhofer/gerpt2	Minixhofer, Benjamin, et al. WECHSEL: Effective initialization of subword embeddings for cross-lingual transfer of monolingual language models. (2021). DOI: http://dx.doi.org/10.18653/v1/2022.naacl-main.293
mGPT	1.3B	sberbank-ai/mGPT	Apache 2.0	https://github.com/ai-forever/mgpt	Shliazhko, Oleh, et al. mgpt: Few-shot learners go multilingual. (2022). DOI: https://doi.org/10.48550/arXiv.2204.07580
MT5	300M 580M 1.2B 3.7B 13B	google/mt5-small google/mt5-base google/mt5-large google/mt5-xl google/mt5-xxl	Apache 2.0	https://github.com/google-research/multilingual-t5	Xue, Linting, et al. mT5: A massively multilingual pre-trained text-to-text transformer. (2020). DOI: https://doi.org/10.18653/v1/2021.naacl-main.41