SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2402.11417 · NAACL · 2024

LoRETTA: Low-Rank Economic Tensor-Train Adaptation for Ultra-Low-Parameter Fine-Tuning of Large Language Models

Zheng Zhang, Yifan Yang, Ngai Wong, Jiajun Zhou

arXiv · PDF · Open in the Atlas

Code that ran

We lifted 8 functions out of this paper's own repositories and ran 4 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.

RepositoryRoleRan
yifanycc/loretta canonical 4 of 8
FunctionStatusWhere it lives
build_message Ran yifanycc/loretta/large_models/message.py
pointer only (licence: GPL-3.0) · get_code("665af05400a43244")
find_module Ran yifanycc/loretta/large_models/lora.py
pointer only (licence: GPL-3.0) · get_code("b5adfd6ac99a1b18")
get_parameter_number Ran yifanycc/loretta/bert_model/run_glue_v5.py
pointer only (licence: GPL-3.0) · get_code("918fc46b5088d174")
normalize_answer Ran yifanycc/loretta/bert_model/metrics.py
pointer only (licence: GPL-3.0) · get_code("e7e75981cb464788")
attn_forward_hook Not yet run yifanycc/loretta/large_models/prefix.py
pointer only (licence: GPL-3.0) · get_code("bdb460cb247501dc")
calculate_metric Not yet run yifanycc/loretta/bert_model/metrics.py
pointer only (licence: GPL-3.0) · get_code("84c658f2c9252878")
f1 Not yet run yifanycc/loretta/bert_model/metrics.py
pointer only (licence: GPL-3.0) · get_code("b989d4bce26f77ce")
prepare_inputs_for_generation Not yet run yifanycc/loretta/large_models/prefix.py
pointer only (licence: GPL-3.0) · get_code("84920a8095dca4fa")

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Various parameter-efficient fine-tuning (PEFT) techniques have been proposed to enable computationally efficient fine-tuning while maintaining model performance. However, existing PEFT methods are still limited by the growing number of trainable parameters with the rapid deployment of Large Language Models (LLMs). To address this challenge, we present LoRETTA, an ultra-parameter-efficient framework that significantly reduces trainable parameters through tensor-train decomposition. Specifically, we propose two methods, named LoRETTA adp and LoRETTA rep . The former employs tensorized adapters, offering a high-performance yet lightweight approach for the fine-tuning of LLMs. The latter emphasizes fine-tuning via weight reparameterization with a set of small tensor factors. LoRETTA achieves comparable or better performance than most widely used PEFT methods with up to 100× fewer parameters on the LLaMA-2-7B models. Furthermore, empirical results demonstrate that the proposed methods exhibit remarkable anti-overfitting capability, effectively improve training efficiency, and enjoy better multi-task learning performance. Plug-andplay loretta library built upon the Huggingface framework and PEFT library are provided. ‡

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2402.11417")
get_code_for_paper("2402.11417")
have("2402.11417")

Connect an agent — have() is free.