- NSE minute-level data for stocks.
- Data comprises opening price, highest price, closing price, traded volume for a stock at every minute in the period from 2018 to 2021.
- Prioritise Building the MLOps pipeline, don’t spend much time in modelling efficiency
for the first iteration of the model (v0), use data in StockAnalyticaData/v0 data folder comprising of these stocks:
- AARTIIND
- ABCAPITAL
For the second iteration of the model (v1), add StockAnalyticaData/v1 data to StockAnalyticaData/v0 - the data from v1 data folder comprising of these stocks:
-
ABFRL
-
ADANIENT
-
ADANIGAS
-
refer to data_processing_reference.ipynb
1. Private Git Repository URL ( 🔗 Tutorial)
- Repository name format:
<ROLLNUMBER_IN_CAPITAL_CASE>_IITMBS_MLOPS_OPPE1 - Grant collaborator access to:
- IITMBSMLOps or
da5014_1@study.iitm.ac.in
- Commit all required files directly to the Git repo - (No separate ZIP file submission)
- Commit history will be scrutinized
- Commit and push to remote after every successful integration with clear comments/commit messages
- ✅ Repository must contain:
- Code/scripts used to complete the objective (
*.py,*.ipynb,*.sh, etc.) - Output files (if any) showing successful completion
README.mdexplaining the purpose of each file- Standard dataset splits used for training
- Any binary artifacts (pickle files, trained models, etc.)
- Code/scripts used to complete the objective (
- ❌ Should not contain:
- Video Screencast
✅ Must cover the following:
- Explanation of the problem statement
- Approach to reach the objective
- Demonstration of cloud compute setup configuration
- Explanation of input files/data
- Demonstration of sequence of actions performed
- Examples of Actions:
- Creating a Virtual Machine
- SSH into VM
- Running scripts
- Detailed explanation of scripts/code and objectives
- Errors encountered & how you resolved them
- Working demonstration in GCP environment
- Explanation of output files/data
3. AI Tool Usage Document - 🔗Reference Document
Should contain:
- AI tools used
- Prompts used
- Link(s) to shared chat(s)
- This can also be included in the Git repo as:
AI_USAGE_DOC.md
- Usage of AI tools is permitted
- Do not discuss anything with your peers.
- Plagiarism if found, will be dealt with severe consequences according to IIT-Madras Student Conduct
- Please submit within allotted time even if the pipeline is incomplete - partial marks are awarded appropriately
- Students are instructed to start the screencast recording only after the completion of whole pipeline or when they are done with their attempts in case of partial completion
- Students are expected to take permission from the proctor about the screencast recording - helps proctor not to flag the instance as malpractice
- Students are expected to inform proctor after successful google form submission
- Students are expected to leave the meet only after completion of scheduled examination time slot - in case of early submissions
- Proctors do note the time of initiating video screencast creation, form submission for operational co-ordination based on student interaction
- In case of no acknowledgement from proctor, please message in Google Meet Chat ( Starting the Screencast Recording or Submission Successful)