How ibm-granite/granite-tsfm works
TSFM is a public time-series foundation-model library with companion inference and fine-tuning services, examples, and demos. A typical forecasting flow takes tabular time-series input through preprocessing and dataset construction into a forecasting pipeline and model, producing forecasts. The toolkit also supports anomaly detection, classification, imputation, conformal methods, and model utilities; separate service paths load models for inference or tuning. Edges are limited to relationships visible in the supplied excerpts, SOURCE INDEX references, or README; many notebook and model-family details were not sampled.
Main entry points: time_series_forecasting_pipeline.py Forecasting Pipeline · dataset.py Dataset Builders · tsfminference/service_handler.py Inference Service API · tsfmfinetuning/service_handler.py Fine-Tuning Service API
Components 26
Foundation Models
- TinyTimeMixer
tsfm_public/models/tinytimemixer/modeling_tinytimemixer.py - TTM Configuration
tsfm_public/models/tinytimemixer/configuration_tinytimemixer.py - TSPulse
tsfm_public/models/tspulse/modeling_tspulse.py - TSPulse Configuration
tsfm_public/models/tspulse/configuration_tspulse.py - PatchTST-FM Components
tsfm_public/models/patchtst_fm/basic.py - FlowState
tsfm_public/models/flowstate/modeling_flowstate.py
Time-Series Toolkit
- Time-Series Preprocessor
tsfm_public/toolkit/time_series_preprocessor.py - Dataset Builders
tsfm_public/toolkit/dataset.py - Forecasting Pipeline
tsfm_public/toolkit/time_series_forecasting_pipeline.py - Anomaly Pipeline
tsfm_public/toolkit/time_series_anomaly_detection_pipeline.py - Classification Pipeline
tsfm_public/toolkit/time_series_classification_pipeline.py - Imputation Pipeline
tsfm_public/toolkit/time_series_imputation_pipeline.py - Conformal Processing
tsfm_public/toolkit/conformal.py - Model Loading
tsfm_public/toolkit/get_model.py - Service Model Config
tsfm_public/toolkit/tsfm_config.py
Inference Service
- Inference Service API
services/inference/tsfminference/service_handler.py - Inference Payloads
services/inference/tsfminference/inference_payloads.py - Forecasting Inference Handler
services/inference/tsfminference/tsfm_inference_handler.py
Fine-Tuning Service
- Fine-Tuning Service API
services/finetuning/tsfmfinetuning/service_handler.py - Tuning Payloads
services/finetuning/tsfmfinetuning/inference_payloads.py - Forecasting Tuning Handler
services/finetuning/tsfmfinetuning/tsfm_tuning_handler.py
Examples and Demos
- Adaptive Conformal TSAD
notebooks/hfdemo/adaptive_conformal_tsad/main_acas_w1.py - TTM Benchmarking
notebooks/hfdemo/tinytimemixer/full_benchmarking/ttm_full_benchmarking.py - NeurIPS Demo
tsfmhfdemos/neurips/app.py
Other components
- Data User
- Hugging Face Models
Connections 33
- Data User → Forecasting Pipeline: submits series
- Forecasting Pipeline → Time-Series Preprocessor: uses preprocessing
tsfm_public/toolkit/time_series_forecasting_pipeline.py - Forecasting Pipeline → Dataset Builders: builds batches
tsfm_public/toolkit/time_series_forecasting_pipeline.py - Forecasting Pipeline → Conformal Processing: supports calibration
tsfm_public/toolkit/time_series_forecasting_pipeline.py - Anomaly Pipeline → Forecasting Pipeline: uses forecasting
tsfm_public/toolkit/time_series_anomaly_detection_pipeline.py - Anomaly Pipeline → Dataset Builders: uses datasets
tsfm_public/toolkit/time_series_anomaly_detection_pipeline.py - Classification Pipeline → Dataset Builders: uses datasets
tsfm_public/toolkit/time_series_classification_pipeline.py - Classification Pipeline → Forecasting Pipeline: reuses pipeline
tsfm_public/toolkit/time_series_classification_pipeline.py - Imputation Pipeline → Dataset Builders: uses datasets
tsfm_public/toolkit/time_series_imputation_pipeline.py - Imputation Pipeline → Forecasting Pipeline: reuses pipeline
tsfm_public/toolkit/time_series_imputation_pipeline.py - TinyTimeMixer → TTM Configuration: uses config
tsfm_public/models/tinytimemixer/modeling_tinytimemixer.py - TSPulse → TSPulse Configuration: uses config
tsfm_public/models/tspulse/modeling_tspulse.py - Hugging Face Models → TinyTimeMixer: provides pretrained models
README.md - Hugging Face Models → TSPulse: provides pretrained models
README.md - Hugging Face Models → PatchTST-FM Components: provides pretrained models
README.md - Hugging Face Models → FlowState: provides pretrained models
README.md - Data User → Inference Service API: requests inference
- Inference Service API → Inference Payloads: validates payload
services/inference/tsfminference/service_handler.py - Inference Service API → Forecasting Inference Handler: dispatches inference
services/inference/tsfminference/service_handler.py - Inference Service API → Service Model Config: loads configuration
services/inference/tsfminference/service_handler.py - Forecasting Inference Handler → Time-Series Preprocessor: constructs preprocessor
services/inference/tsfminference/tsfm_inference_handler.py - Forecasting Inference Handler → Conformal Processing: uses calibration
services/inference/tsfminference/tsfm_inference_handler.py - Data User → Fine-Tuning Service API: requests tuning
- Fine-Tuning Service API → Tuning Payloads: uses payload types
services/finetuning/tsfmfinetuning/service_handler.py - Fine-Tuning Service API → Forecasting Tuning Handler: dispatches tuning
services/finetuning/tsfmfinetuning/service_handler.py - Fine-Tuning Service API → Service Model Config: loads configuration
services/finetuning/tsfmfinetuning/service_handler.py - Forecasting Tuning Handler → Time-Series Preprocessor: uses preprocessing
services/finetuning/tsfmfinetuning/tsfm_tuning_handler.py - Forecasting Tuning Handler → Dataset Builders: builds training data
services/finetuning/tsfmfinetuning/tsfm_tuning_handler.py - Forecasting Tuning Handler → Model Loading: uses learning-rate finder
services/finetuning/tsfmfinetuning/tsfm_tuning_handler.py - Data User → Adaptive Conformal TSAD: runs example
README.md - Data User → TTM Benchmarking: runs benchmark
README.md - Data User → NeurIPS Demo: uses demo
README.md - NeurIPS Demo → Forecasting Pipeline: uses forecasting pipeline