update readme

README.md

@@ -20,12 +20,14 @@ The official code for ICLR 2024 paper: "TEMPO: Prompt-based Generative Pre-train
 
 TEMPO is one of the very first open source Time Series Foundation Models for forecasting task v1.0 version.
 
-<div align="center"><img src=pics/TEMPO.png width=80% /></div>
+![TEMPO-architecture](pics/TEMPO.png)
+
 
 
 Please try our foundation model demo [[here]](https://4171a8a7484b3e9148.gradio.live).
 
-<div align="center"><img src=pics/TEMPO_demo.png width=80% /></div>
+![TEMPO-demo](pics/TEMPO_demo.png)
+
 
 # Build the environment
 
@@ -57,8 +59,7 @@ After training, we can test TEMPO model under the zero-shot setting:
 ```
 bash [ecl, etth1, etth2, ettm1, ettm2, traffic, weather]_test.sh
 ```
-
-<div align="center"><img src=pics/results.jpg width=90% /></div>
+![TEMPO-results](pics/results.jpg)
 
 
 # Pre-trained Models
@@ -69,16 +70,18 @@ You can download the pre-trained model from [[Google Drive]](https://drive.googl
 
 Here is the prompts use to generate the coresponding textual informaton of time series via [[OPENAI ChatGPT-3.5 API]](https://platform.openai.com/docs/guides/text-generation)
 
-<div align="center"><img src=pics/TETS_prompt.png width=80% /></div>
+![TEMPO-prompt](pics/TETS_prompt.png)
+
 
 The time series data are come from [[S&P 500]](https://www.spglobal.com/spdji/en/indices/equity/sp-500/#overview). Here is the EBITDA case for one company from the dataset:
 
+![Company1_ebitda_summary](pics/Company1_ebitda_summary.png)
 
-<div align="center"><img src=pics/Company1_ebitda_summary.png width=80% /></div>
 
 Example of generated contextual information for the Company marked above:
 
-<div align="center"><img src=pics/Company1_ebitda_summary_words.jpg width=80% /></div>
+![Company1_ebitda_summary_words.jpg](pics/Company1_ebitda_summary_words.jpg.png)
+