OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
AI Models

Time LLM

Time LLM is an open-source Python research implementation for time-series forecasting by reprogramming large language models. It converts time-series inputs into text-prototype representations and augments their context with prompts such as domain knowledge and task instructions, while keeping the language-model backbone intact.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Time LLM — OpenViST project profile in AI Models.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use it when you want to investigate or run the repository's Python experiments for adapting language models to time-series forecasting, rather than access a hosted forecasting tool.

Who it is for

  • Researchers exploring language models for time-series forecasting.
  • Practitioners comparing prompt-guided forecasting methods or adapting language-model backbones to time-series inputs.

Documented uses

  • Set up and examine the implementation's forecasting experiments.
  • Evaluate the provided examples on preprocessed ETT datasets.
  • Study how input representations and declarative prompts shape language-model forecasting.
  • Compare documented backbone choices, including Llama-7B, GPT-2 and BERT.
Before you choose

Requirements & limits

The repository is publicly available under the Apache-2.0 license.

  • The documented environment specifies Python 3.11 via MiniConda and a pinned dependency list, including PyTorch, Transformers, PEFT and DeepSpeed.
  • The README provides preprocessed dataset access and example scripts for ETT evaluations.
  • Detailed hyperparameter descriptions are directed to run_main.py, run_m4.py and run_pretrain.py.
  • The maintainers state that major updates to the main codebase are not planned, though constructive pull requests are welcome.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Evaluate Time-LLM on an ETT dataset using an example script

Run one of the repository's documented ETT example experiment scripts after preparing its dependencies and datasets.

Before you start

  • Python 3.11 from MiniConda and the dependencies installed as described in the repository README
  • The downloaded dataset contents placed under ./dataset

Steps

  1. Choose one documented ETT example script, such as ./scripts/TimeLLM_ETTh1.sh.

  2. Run it from the repository using the command below.

    Documented command
    bash ./scripts/TimeLLM_ETTh1.sh

  3. For other documented ETT examples, use the corresponding ETTh2, ETTm1, or ETTm2 script.

Expected outcome

The selected ETT experiment script is invoked to tune and evaluate the model.

Guide limitations

  • The README identifies these as demonstration scripts but does not specify expected metrics or guarantee successful execution.
  • Hyperparameter descriptions are referenced in run_main.py, run_m4.py, and run_pretrain.py.
  • The repository lists no major planned updates to its main codebase.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
10 October 2026
Original website title
KimMeen/Time-LLM · Official homepage title · retained in its source language
How we review sourcesSuggest a correction