To extract information from the images of the climate reports

[rkclimate20]

• Add readme for the pipeline

[Avika2211]

https://a269241f-175f-459c-b826-9747e0a38359-00-12e0lvwqle8y6.kirk.replit.dev/

hosted the tool on replit

[Avika2211]

Issues faced so far:
currently my tool is integrated with a hugging face model but for more specific, detailed and accurate classification without hallucination, i need to implement supervised learning. The major issue being faced in this process is the procurement of a good dataset which suits our use case. if i go and train the model manually and procure each and every type of figure, it'll take extreme amount of time and effort. I wonder if there is a better way to do it.