Skip to content
View Kaustubh1235's full-sized avatar
🌴
On vacation
🌴
On vacation

Block or report Kaustubh1235

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Kaustubh1235/README.md

banner

typing


πŸ† Achievements unlocked

stanza reef tvm

goclaw mex SWE-bench

Each card is a repo that took my code. Click one to open the PR.

6
PRs merged upstream
5
repos that took my code
+1,563
lines merged
27k β˜…
combined stars of those repos
2
in review (Apache TVM, SWE-bench)

🎯 Quest log β€” merged

πŸ… Project What I shipped
πŸ₯‡ stanfordnlp/stanza 7.9k β˜… #1693 Bounded the lemmatizer's results cache (LRU) and made it the default, so long-running pipelines stop growing without limit
πŸ₯‡ #1692 Kept the cache off words a contextual lemmatizer decides, so her β†’ she / her is still resolved by context after it's been seen
πŸ₯ˆ nextlevelbuilder/goclaw 3.6k β˜… #1583 Fixed agent deletion failing on an orphaned-vault path collision β€” migration, widened unique index, integration test that reproduces it
πŸ₯ˆ Human-Agent-Society/reef 7.6k β˜… #553 Harness requires kinds, so a program can declare the binary a release names
πŸ₯‰ #714 Tell the model when a grounding document was cut short instead of ending mid-sentence silently
πŸ₯‰ mex-memory/mex 1.8k β˜… #196 Stopped targeted graph reads from re-auditing the whole store

βš”οΈ Current quest

apache/tvm #20617 β€” ConvertLayout segfaulted on add(x, float32(1)): layout inference dereferenced a null cast on the PrimValue operand. Reproduced from a no-LLVM build β†’ fixed β†’ regression test β†’ all 5 CI jobs green. Awaiting review. SWE-bench #670 β€” run reports now count only the instances you selected.


πŸ“Š Stats

stats languages

streak

contribution chart


🧰 Loadout

stack


πŸ› οΈ Side quests

πŸŽ™οΈ Vocat
Real-time voice conversation at sub-second latency.
WebRTC β†’ VAD β†’ Whisper β†’ streaming GPT-4 β†’ Redis β†’ ElevenLabs β†’ WebRTC.
πŸ“ˆ AIEquity Traders
Multi-agent trading-floor simulation. Autonomous agents talking to local and remote MCP servers.
🧬 BioGPT hallmark classifier
Small biomedical language model fine-tuned for hallmark-of-cancer classification.
🌬️ ERA5 wind-speed LSTM
Wind-speed prediction over time and altitude from ERA5 reanalysis data.
πŸ›‘οΈ NETOPTAI
Phishing-site detection with the MLOps around it: MongoDB ingestion, training pipeline, deployment.
βš–οΈ Legally
LLM legal assistant for Indian law, for people who can't afford a lawyer for a first question.

contribution snake

footer

Pinned Loading

  1. vocat vocat Public

    Python

  2. era5_project era5_project Public

    LSTM WIND SPEED PREDICTION USING ERA5 DATASET

    HTML

  3. 1d-lstm 1d-lstm Public

    smart irrigation

    Jupyter Notebook

  4. 3d_pose 3d_pose Public

  5. BIO-SMALL-LANGUAGE-MODEL BIO-SMALL-LANGUAGE-MODEL Public

    Jupyter Notebook

  6. MCP MCP Public

    Python