link to Granite
Browse files
README.md
CHANGED
|
@@ -14,6 +14,9 @@ tags:
|
|
| 14 |
|
| 15 |
## Model Description
|
| 16 |
|
|
|
|
|
|
|
|
|
|
| 17 |
[PatchTST](https://arxiv.org/abs/2211.14730) was originally released prior to the interest in creating pre-trained, zero-shot time series foundation models that were capable of state-of-the-art performance on out of sample datasets. PatchTST-FM (patched time-series transformer-based foundation model) essentially has the architectural simplicity of PatchTST, but differs in some crucial ways. Coupled with a revised training strategy and a significantly larger training corpus, we are able to train a model that achieves state-of-the-art results on [GiftEval](https://huggingface.co/spaces/Salesforce/GIFT-Eval).
|
| 18 |
|
| 19 |
The architecture incorporates the following changes:
|
|
|
|
| 14 |
|
| 15 |
## Model Description
|
| 16 |
|
| 17 |
+
|
| 18 |
+
This model card is for the non-commercial, research version of PatchTST-FM-r1. Please also check-out the Apache-2.0 licensed [IBM Granite version](https://huggingface.co/ibm-granite/granite-timeseries-patchtst-fm-r1).
|
| 19 |
+
|
| 20 |
[PatchTST](https://arxiv.org/abs/2211.14730) was originally released prior to the interest in creating pre-trained, zero-shot time series foundation models that were capable of state-of-the-art performance on out of sample datasets. PatchTST-FM (patched time-series transformer-based foundation model) essentially has the architectural simplicity of PatchTST, but differs in some crucial ways. Coupled with a revised training strategy and a significantly larger training corpus, we are able to train a model that achieves state-of-the-art results on [GiftEval](https://huggingface.co/spaces/Salesforce/GIFT-Eval).
|
| 21 |
|
| 22 |
The architecture incorporates the following changes:
|