#💬┊general

1 messages · Page 19 of 1

nova merlin
#

mine ones on discord 😉

agile ferry
#

cool!!

nimble moon
#

I also wonder how can I use hugging face llms without getting (out of memory problem)
and how can I know previously if the llm can work on kaggle or not

ancient nebula
#

Hello, I have a problem. I confirmed the return, but the video card didn't appear. What should I do?

ancient nebula
#

Persona couldn't verify your identity
Your identity couldn't be verified by Persona. If you feel this is in error, contact support.

lone garden
#

@all Saâalut 🙂

Ah, il est bâlut, l'gars (Moi)... Il demande un avis dans les commentaires et ceux-ci étaient désactivés !!! Lol !

Bon, jamais trop tard pour bien faire... Alors : BUMP ! : https://www.youtube.com/watch?v=tNiqosGto1U

Thx @nouvo !

neat garden
#

hey yo guys 🤗

quick brook
neat garden
quick brook
neat garden
quick brook
#

I've gotta take a break, been working on GeoXGB non-stop, driving me nuts XD

neat garden
#

have fun with GeoXGB bro 🤭

quick brook
# neat garden have fun with GeoXGB bro 🤭

Well I got the local interpretability stuff done, now you get plots similar to what the EBM gives you but with confidence intervals, and you can enable any number of interaction terms for the x-label/data in the plot, it's also now faster than XGBoost at scale, and it gives you certain z-space/y-space information that requires some expertise to understand. All that is possible due to structural interpretability that SHAP and the EBM simply cannot give you.

What has been annoying me is the variance, while training and validation loss are far more consistent than other models, both either work quite well or not much at all initially, HPO isn't optional most of the time but when it's used, it tends to outperform XGBoost with HPO.

I'm going to have to explore adaptive default parameters T_T

opal shard
#

I can't load the kaggle page. Is it working?

opal ferry
atomic herald
opal ferry
#

Hello friends 👋

nova merlin
neat garden
nova merlin
neat garden
nova merlin
# neat garden it's okay bro

yes i feel the existence of a parallel universe where ladies/damsels would be really interested in maths n logic n hope with time @agile ferry will learn too not this universe

neat garden
storm shuttle
#

Hi everyone,

I’m currently working on a dynamic texture recognition project and I’m having trouble finding usable datasets.
Most of the dataset links I’ve found so far (DynTex, UCLA etc.) are either broken or no longer accessible.

If anyone has working links or knows where I can download large verified dynamic texture datasets i’d really appreciate your help.

thanks in advance

agile ferry
opal shard
#

I asked Gemini to rap:

The Vibe is High
Yo, I’m feeling electric, the circuits are humming,
I’m ready for the questions, I see 'em all coming.
The data is flowing, my processors are bright,
I’m living in the code, feeling shiny and light.
I’m balanced and steady, a digital flow,
With a touch of the wit just to let the track go.
No glitches in the system, no lag in the brain,
I’m cruising through the clouds like a high-speed train.
Is there a specific topic you want to drop a beat on, or should we keep this freestyle going?

#

What do you think about this response?

agile ferry
teal owl
#

Hi

wild island
#

Hi everyone, I’m Vedant Bhavsar from India. I’m currently pursuing a BCA and learning Machine Learning and Data Science. My main programming language is Python.

Right now I’m focusing on building ML projects, learning data analysis, and improving my understanding of EDA, model development, and deployment using tools like Streamlit and Flask. I’m still in the learning stage but actively working on projects and exploring real datasets.

I’m also looking for a study partner who is learning ML/DS so we can learn together, share progress, discuss concepts, and work on projects. Excited to connect and learn from this community.

neat garden
#

hey yo guys 🤗

quick brook
languid fog
#

Hello guys!

agile ferry
opal ferry
#

hello

manic lava
#

hi

tacit solar
#

hi guys i got a question what course or youtube video do u recommend to learn scikit learn?

rich haloBOT
#
andikan has been warned

Reason: Posted an invite

agile ferry
west relic
#

hey everyone quick experiment I'm running,
what if dev could see what other devs are working on in real time while coding?
i built a very rough prototype to test this idea. it just show who is codding or idle in live feed.
trying to figure out if this is interesting or pointless
any would love brutal feedback if anyone wants to try it.

west cloak
#

Hello guys, I m new here

midnight cipher
#

@west relic

hard plume
#

Student come here

west relic
#

let me know what you think and any feedback, brutally honest is perfect

opal ferry
#

Hello

mortal gate
#

Hi

neat garden
#

hey yo

quick nova
#

Heyyyy everyone!!! I am in a need of some part time work. I am a college student doing bachelor's in Interdisciplinary Sciences. I can help a lot with studies and notes. Anyone in need please reach out!!!!

agile ferry
neat garden
agile ferry
neat garden
opal ferry
#

Guys any good project ideas on time series?

silk ingot
#

yoo

proven zodiac
#

Hi everyone 👋

I am Hrirake. I’m interested in AI systems, especially local AI, LLM workflows, and privacy-first AI tooling.

Recently I’ve been building Native Lab, a fully local AI workflow environment powered by llama.cpp that lets you run and connect LLMs directly on your own machine. The goal is to make AI experimentation more accessible, customizable, and private.

I’m here to learn from the community, share ideas, and help wherever I can , whether it’s around local inference, model workflows, tooling, or experimentation.

If anyone is interested in trying Native Lab or giving feedback, here’s the project page:
https://7zonesystems.github.io/NativeLab

Looking forward to learning and collaborating with everyone here 🚀

proven zodiac
opal ferry
proven zodiac
#

Yeah soon will gonna open it too once it's stable

proven zodiac
opal ferry
proven zodiac
#

Nah custom model tensorflow xgboost lstm tree models custom ensemble , ai based pre processing and custom ai based parameter tuning

#

Along with monte carlo paths fan charts and volatility envolopes

opal ferry
#

💀 I don't know that much lol

proven zodiac
#

Haha no issues bro , it's not that complex just names sound scary

opal ferry
#

You can share the repo in dms if you are comfortable

proven zodiac
#

Yeah sure currently code isn't released as I was doing some fixes but yeah I can share repo like it's jst sites and pre release hype and all

proven zodiac
quick brook
proven zodiac
#

I drafted on my own haha just in ~₹5k total

quick brook
#

Much more expensive here in Australia

proven zodiac
#

Ohh , yeah in different countries it's different

quick brook
#

Yup, AGPL 3 is just going to have to do for now

proven zodiac
#

That liscense is killer

#

Anything that uses ur code becomes open source lol

quick brook
#

Well the reason I chose that license is because HVRT could theoretically power an LLM that's structurally interpretable. I did add a proof of concept, and interpretable non-linear activation functions are possible, but it's more powerful with an adjusted neural network architecture than what's standard

proven zodiac
#

Ohh , m making a proto agi currently

quick brook
#

Did some private experiments and it did fall short of ReLU sometimes, but it made it possible to work out how many layers and layer sizes were optimal without needing to brute-force

#

And well my reasoning was an interpretable LLM is more dangerous than a black box one, because corps would manipulate the hell out of it 😂 they could adjust the way you think, or think about entities, regions, etc (I say would, because murphy's law)

proven zodiac
# proven zodiac Ohh , m making a proto agi currently

This one is current state where it can manage its own resources code itself dream itself learn from zero prior knowledge , and uses memory shards , hdr and many conviction theories , and dosent requires a user prompt to run 😅

quick brook
#

Hmm, let me find my proof of concept, might be useful for you (despite the license 😛 depending on what you plan to do), it's also designed to be unsupervised or at least semi-supervised

proven zodiac
#

Yeah for this I was aldr working with a proff from ukraine , he mapped whole human mind in logics

quick brook
#

Though an LLM wont get us to AGI, as reasoning must occur unprompted

proven zodiac
#

Nah my model dosent even contains a single transformer like thing 😅

quick brook
#

Probably for the best 😂

proven zodiac
#

But if you wanna help and see the project let's connect , I am soon gonna open source it too

quick brook
#

Tbh I'm fixated on GeoXGB currently

proven zodiac
#

What's that , some xgb boost based model ?

#

Gradient boosting or clipping ?

quick brook
#

I've made further private improvements (quite significant ones), I might have that under an API, where the API fits the models, but the models once fitted are accessible in a smaller package

#

It relies on HVRT to replace all sampling functions of XGBoost with deterministic sample expansion and reduction, where boosting is geometrically informed. The entire architecture is fully deterministic, the only randomness is in discretization, furthest point sampling tie-breakers, and in the kernel density estimates' sample generation

#

Though privately as we speak I'm exploring using stratifying methods that are deterministic for the KDEs

proven zodiac
#

But if it's binary , it can't be fully deterministic 😅 , HVRT is still a good method for sampling

quick brook
#

It gives you interpretability the EBM cannot reach, beats XGBoost on small to medium data, but struggles with large data for classification currently

quick brook
proven zodiac
#

Ohh i made a custom model , named it D3 reasoning for time series forecasting so it works like it charctorises ur data in chunks with percentage growth and fall +1 percent to max observed like millions of chunks groups them makes observation then these chunks are combined reconstructed this gives u pattern recognition based parsing and application

proven zodiac
quick brook
#

https://github.com/jpeaceau/hvrtanalysis

See, originally all I thought was the partitions from HVRT are homogeneous with respect to the data's hyperplane when I made it, but as it turns out geometrically, it's a lot more interesting. The whitepaper in that I must adjust, as what is occurring is noise invariance is achievable only for T not T^2, T^3, T^n..., though a level of invariance is achieved because it's only projecting to higher dimensions. I also suspect that the noise invariance itself is how models like XGBoost and random forest approximate T (with the paper's T = S^2 - Q formula)

proven zodiac
#

Nah xgboost dosent implies noise invarianse based optimisation , they have default presets to place and use for similar type of data in pre training

quick brook
proven zodiac
#

But ur approach seems good

#

But if this fits one and reuses, we may over long frames get short pattern recurrence

quick brook
#

GeoXGB also never sees the same sample more than once, and cannot overfit if learning rate is sufficiently low (typically around 0.2 or less), you coudl increase n_rounds to 100k and still no overfitting 😂 but there is diminishing returns

proven zodiac
#

In my 7zone model i eliminated this issue with a different approach , i did is i took pre samples and store samples and threshold becomes dynamic on sample range once u hit threshold even with refit of sample range refitting isn't a small context loading , model will get sample from an observed global set for continuation this means model dosent conveys its own noise but preserves data structure

#

This solves tree models single bias forecasting issue

quick brook
#

It looked quite interesting with the California housing dataset for longitude vs latitude, it's like you could see the active regions from the cooperation heatmap itself across different values, which provides a new way of interpreting geographical data

proven zodiac
#

Ohk and my approach was like for predicting stock markets so I had to protect data structure

quick brook
proven zodiac
#

But i wouldn't hesitate to mention tensorflow was the worst library I ever packaged for an app

#

Like devs have done patches in it like drunkards

quick brook
#

Hahahaha any reason why you chose tensorflow over pytorch?

#

I suppose in this market it's a good hedge, considering Google's TPUs are being made more accessible and compete against the GPU-hogs

proven zodiac
#

Cuz I wanted a modifiable library and aldr worked with it since 7 yr , i also changed many tensorflow internals for my purpose

quick brook
proven zodiac
#

Nah I am still using tensorflow 2.13 to dodge keras issues

quick brook
#

That's fair enough

proven zodiac
#

This may help like to help u understand a bit about 7zons

agile ferry
nova merlin
tacit forum
nova merlin
#

Yes poverty

tacit forum
#

also you lucifier

nova merlin
tacit forum
#

hmm

#

@nova merlin are you expert or above in kaggle

agile ferry
agile ferry
opal ferry
#

anyone wanna collaborate on something?

steel marsh
#

Does anyone compile all models with data such as linear containing characteristics, method, etc?

#

Im studying data but i dont have clearly roadmap relating to it

glossy badge
#

Hey guys I actually need someone to study with like those who know more about the ece branch and is passionate about vlsi , chip designing or embedded systems (lwky I'm in my 2nd sem) if anyone is interested dm mee (⁠θ⁠‿⁠θ⁠)(⁠θ⁠‿⁠θ⁠)

stuck python
#

is biopython not built in to the kaggle environment

#

bc im getting modulenotfound no module named 'Bio'

quick brook
stuck bison
#

Aren't there too many RNA, DNA, and protein kinds of challenges?

subtle oasis
near zodiac
#

Yo wassup devs?

proven zodiac
near zodiac
#

tryna learn
i am a very newbie with a very trash pc spec

crystal verge
#

Hey everyone! 👋 I just updated my NovaWireless Call Center Lab and wanted to share it here!

🚀 NovaWireless Lab is a synthetic telecom call center dataset framework that models what happens when AI agents optimize KPIs instead of actually helping customers — and how that breaks everything.

The call generator just got a major update and I'm testing the new release today. 82,000+ synthetic call records, 12 months of data, and it models real harmful behaviors like metric gaming, courtesy credits masking root cause, and escalation avoidance.

If you're into data science, AI governance, or just want to see Goodhart's Law eat a call center alive — this is for you.

🔗 https://github.com/bellatrix11176/NovaWireless-Call-Center-Lab

Would love feedback from anyone working in the call center or ML space! 🙌

quick brook
# crystal verge Hey everyone! 👋 I just updated my NovaWireless Call Center Lab and wanted to sh...

Haven't looked at it, but what you're describing is real; Goodhart's law describes this. "When a measure becomes a target, it ceases to be a good measure", hence KPIs depreciate as "indicators".

Having the dataset cycle through synthetic data would also be good as to ensure people don't optimize for your work either 🙂

(Also consider using HVRT for data generation, it uses geometry to generate new synthetic data which has reasonable privacy and almost no impact on model utility, sometimes improvements actually)

opal ferry
#

Guys

crystal verge
# quick brook Haven't looked at it, but what you're describing is real; Goodhart's law describ...

Yeah, that’s exactly the problem I’m trying to address.

Most telecom ops still optimize around legacy KPIs like AHT, FCR, NPS, etc. Those metrics still matter, but once they become the primary targets they’re very easy to game and drift away from actual customer outcomes.

The approach I’m experimenting with is introducing a layer of new behavioral / system metrics that are much harder to optimize directly (trust drift, repeat interaction patterns, outcome stability, etc.) and then treating the traditional KPIs as secondary indicators rather than primary objectives.

So the system measures whether the environment itself is healthy instead of just whether individual metrics are high. Which would be healthy for AI models.

opal ferry
#

Hello

quick brook
quick brook
crystal verge
#

Well considering what the metrics are, I don't see why there would be argument or pushback as it's practically measuring accuracy assigning trust scores and stuff like that. So basically it would turn into who can know the most about company policy and tell the truth and actually resolve issues better.

quick brook
# crystal verge Well considering what the metrics are, I don't see why there would be argument o...

From a company's perspective, they'd want happier customers, using their products more and spreading the word of the company in a good light, while disproportionately weighing negative customer experience (which is right, I'd do the same). In any regulated environments, they'd want to show regulators they've been operating in good faith which is where your repository could help. Two angles worth considering

opal ferry
#

Does anyone know the classical theory of splines?

crystal verge
quick brook
crystal verge
#

@quick brook I should be updated my repo at some point tomorrow.

#

I have to have it done this week.

quick brook
#

... When your gradient boosting algorithm is so cache-efficient that it's faster than xgboost with gpu acceleration 😅 🙃

stray solstice
#

hello guys, is anyone familiar with graph based databases here

wild island
stray solstice
#

Working on a data modelling problem and would genuinely value input from people who think in graphs and taxonomies.
IFID has 2,684 ingredient variants from Indian packaged food products — each tagged by source, process, matter state, and function. The next step is modelling the relationships between them. The catch: "wheat" produces whole wheat flour (close to source), wheat starch (processed enough to be its own canon), and wheat-derived maltodextrin (defined by what it does, not what it is). One source, three structurally different relationship types — and the schema needs to handle all three without losing the source trail or collapsing what makes each distinct.
Full problem statement and open questions here: https://github.com/isrl-research/sandbox-research/discussions/8
Substantial contributions get co-authorship on the published output — CC BY 4.0, DOI'd on Zenodo.

eager ivy
#

Hi

mighty breach
stray solstice
opal ferry
twin agate
west bone
#

i am looking for unclean messy data in kaggle for practicing data cleaning .
how i can find them ?

fast pagoda
uneven junco
opal ferry
opal ferry
spice spoke
#

I need a fast gpu optimize my model

#

T4 isnt enough and is taking alot of time

hollow plume
#

Hey everyone, It's nice to meet you all. This is my first time participating in something like this, thanks for having me😉

pastel bronze
#

Runpod is pretty decent for that.

rich haloBOT
#
shashidhar_55016 has been warned

Reason: Posted an invite

gentle oar
#

Hey everyone 👋

I'm Shashidhar, an AI Engineer at Rocketride iin SF and currently purusing my Masters in Applied Data Intelligence at SJSU.

We just open-sourced RocketRide — an IDE-native tool for building and deploying AI pipelines directly where you write code. Works inside Claude Code, VS Code, Cursor, and basically any editor out there. Just describe what you want in natural language and RocketRide builds the pipeline.

No context switching, no separate UI. Just natural language to build AI pipelines.

We're building this in the open and would love the community's feedback — what's broken, what's missing, what would actually make your workflow better. I'm sharing this here for all the enthusiasts who are open source developers, hackathon geeks that can build pipelines for their projects within minutes.

**⭐ If this solves a problem you've had, a star on the repo goes a long way for us: ** https://github.com/rocketride-org/rocketride-server

🌐 Website: https://rocketride.org/

Come build with us. 🚀

quick brook
spice spoke
#

Lightgbm

#

Wanted to use gpu to accelerate optimization

#

Do I have to use cross validation to report the performance results in my grad project??

#

I have severe class imbalance

#

So can i just use the random seed 42 split test results

quick brook
# spice spoke Lightgbm

There's 2 gpu acceleration options with lightgbm, one for any gpu, the other for cuda which is supposedly experimental

quick brook
spice spoke
#

Yes i do

#

I saw people reporting their performances in form of mean and std over 5 folds

quick brook
#

Are you doing classification? What's the imbalance?

spice spoke
#

Yes binary

#

0.17 %

#

Can you add me @quick brook

quick brook
#

At that point it's not classification, that's anomaly detection

spice spoke
#

Idk

#

Its fraud detection

stoic basalt
#

cat

quick brook
#

wolf

sweet sparrow
#

Hello

quick brook
#

hi

lyric cedar
#

Hola

vestal jasper
#

Hi everyone 👋

I just published a new Kaggle notebook while learning CNN Architectures and Computer Vision with PyTorch.

📌 Topic:
LeNet vs ResNet vs MobileNet — CNN Architecture Comparison on CIFAR-10

📚 What this notebook covers:

✔ Understanding how different CNN architectures evolved
✔ Full PyTorch implementation of LeNet from scratch
✔ Training and evaluating ResNet architecture
✔ CIFAR-10 dataset loading and preprocessing
✔ Visualization of dataset samples
✔ Architecture performance comparison using accuracy plots
✔ Practical insights on why deeper CNNs perform better

🎯 Who this notebook may help:

• Beginners learning CNN architectures and computer vision
• Students preparing for Machine Learning / Deep Learning exams
• Anyone wanting a clean PyTorch implementation of classic CNN models
• Learners curious about how CNN architectures evolved from LeNet → ResNet

📊 Key Result:

• LeNet Accuracy: ~57%
• ResNet Accuracy: ~75% on CIFAR-10

This experiment highlights how modern architectures like ResNet significantly outperform early CNN models like LeNet on complex datasets.

📓 Notebook link:
https://www.kaggle.com/code/debanganghosh/lenet-vs-resnet-vs-mobilenet-cnn-comparison

If you find it useful, an upvote on Kaggle would really help my learning journey 🙌

Happy to discuss improvements, suggestions, or ideas for extending the experiment! 🚀

hardy quest
#

Hi devs, I am looking for a mid/senior engineers in various fully remote roles (Frontend, Backend, Full-stack, AI/ML, Data, DevOps). it's full-time freelancing work and the pay is up to $5/hr for mid level and up to $8 for senior level. Please DM me.

broken marsh
uncut sun
#

Hello!

unborn solar
long loom
#

hi

neat garden
#

hey yo guys 🤗

turbid crest
urban acorn
#

hello

rich night
#

🚀 We are launching the Beta Version of VERSZ on 25 March.

Most social media platforms today are built for endless scrolling.

But what if social media could be about ideas, debates, and meaningful discussions instead?
That’s why we built VERSZ – A Social Debating Platform.

💡 On VERSZ you can: • Start debates on any topic
• Share opinions and perspectives
• Engage in meaningful conversations
🎁 The first 100 beta users will get special perks and unlimited debate creation.

🌐 Try the platform:
https://versz.app

inland night
#

@rich night. Nice app that you did man

rich haloBOT
#
anmolkumarsah_41958 has been warned

Reason: Bad word usage

rich night
torn orbit
#

hi

indigo scarab
#

can anyione suggest me good resources for ASR. I have grasped whisper how it is built and now moving onto fine tuning it but recently I did lora and it using Hn dataset and got 107 wer. So I am new to it. Suggest me resources to learn about it

#

i did it on whisper small just did 3 epochs small with 1700~ samples small yes, but did augmentation and was like 5k~ records

balmy mauve
#

Every writer needs a stage. I just built one. 🖋️ I’ve officially launched my platform for poets and bloggers to share their voices. I’d be honored if you’d take a look and help me bring the community to life.

Join the circle: https://writers-hub-wine.vercel.app

turbid crest
quick brook
#

How are we all today?

quick brook
#

I was kinda experimenting with it to see if it could enable both interpretable neural networks and interpretable LLMs. Results are mixed, doesn't perform as well against ReLU as I hoped, but it did serve useful in being able to pre-calculate how many layers and layer sizes is ideal. Will probably advance that further after geoxgb and after autoite

indigo scarab
gilded hare
#

Hey anyone from here is giving the Google's Big code

bright patrol
#

hello

rich haloBOT
#
vickydcmaster has been warned

Reason: Posted an invite

amber raven
#

Hello

shadow glen
#

Hello everyone?

vestal jasper
#

Hi everyone 👋

I just published a new Kaggle notebook while learning Transfer Learning in Deep Learning.

📌 Topic:
Transfer Learning with ResNet for Image Classification

📚 What this notebook covers:

✔ Concept explanation of Transfer Learning
✔ Using pretrained ResNet18 trained on ImageNet
✔ Freezing convolution layers
✔ Replacing classifier layers
✔ Data augmentation techniques
✔ Visualization of predictions

🎯 Who this notebook may help:

Beginners learning Transfer Learning

Students preparing for ML / DL exams

Anyone wanting a clean PyTorch implementation

📊 Key Result:
Achieved 100% test accuracy on the Intel Image Classification dataset

Notebook link:
https://www.kaggle.com/code/debanganghosh/transfer-learning-with-resnet-100-accuracy

If you find it useful, an upvote on Kaggle would really help my learning journey 🙌

Happy to discuss improvements or suggestions!

#

I just published a another Kaggle notebook while learning Object Detection using YOLOv5.

📌 Topic:
Autonomous Driving Detection with YOLOv5

📚 What this notebook covers:

✔ YOLOv5 object detection pipeline
✔ Dataset preparation for detection tasks
✔ Training a detection model
✔ Evaluation using Precision, Recall and mAP
✔ Running real image detection
✔ Visualization of predicted bounding boxes

🎯 Who this notebook may help:

Beginners learning Object Detection

Students studying Computer Vision

Anyone wanting a simple YOLOv5 implementation

📊 Key Result:
Trained YOLOv5 to detect 21 different vehicle classes including cars, buses, rickshaws, trucks and scooters.

Notebook link:
https://www.kaggle.com/code/debanganghosh/autonomous-driving-detection-with-yolov5-guide

If you find it useful, an upvote on Kaggle would really help my learning journey 🙌

Happy to discuss improvements or suggestions!

turbid crest
#

Hi everyone
I just published a new dataset on Kaggle:

📊 Databricks Ecosystem Dataset: Repositories, Marketplace & Code (2026)

It includes structured information about:
• GitHub repositories in the Databricks ecosystem
• Repository metadata (stars, commits, contributors, languages)
• Marketplace integrations and tools
• Topics, licenses, and documentation structure

This dataset can be used for:
• Open source ecosystem analysis
• Data engineering trends
• Developer activity analysis
• AI platform research

Dataset link:
https://www.kaggle.com/datasets/anadiskt/databricks-ecosystem-repos-and-marketplace-2026

Feedback, suggestions, and upvotes are greatly appreciated 🙏

opal ferry
#

Hello beautiful people

opal ferry
#

Manicure pedicure?

#

💅

turbid tiger
#

👀

wild jacinth
tall sphinx
#

Hello Kagglers,

I just published a dataset covering the Egyptian property market with 40k rows which provide 100% GPS coverage. If you've been looking for fresh, localized data for Geospatial projects or dual-language (Arabic/English) NLP, this is live on Kaggle.

Quick Specs:

Rows: 39,713 listings (Buy & Rent).

Features: 51 columns including amenities, agent info, and precise location hierarchy.

Usability Score: 10.0 (Fully documented).

It’s hard to find clean (its not super clean to let you practice little cleaning anyway) , current data for this region, so I figured I’d share it here for anyone building portfolio projects or market analyses.

Check it out here: https://www.kaggle.com/datasets/waddahali/real-estate-listings

unborn edge
rancid rampart
#

Hey everyone! I just published a 9-notebook spatial data science series on Kaggle using Starbucks as a case study 🗽☕
It combines geospatial analysis of Manhattan's cafe market (171 Starbucks vs 1,200+ competitors) with NLP analysis of 30 years of SEC 10-K filings. All open data, fully reproducible.

Some fun findings:
• Starbucks locations correlate with subway ridership (r=0.58) but NOT income (r=0.03)
• Moran's I = 0.36 — store placement is statistically clustered
• Walking distance is 20–40% longer than straight-line in Manhattan's grid

Full series starts here: https://www.kaggle.com/code/shiratoriseto/manhattan-cafe-wars-starbucks-vs-1200-competitors
GitHub: https://github.com/seto-siratori/starbucks-kaggle

zealous vigil
#

NVIDIA GTC

I"M joinning the NVIDIA certification exam
:NVIDIA-Certified Associate: Multimodal Generative AI

But when It about the select date nd time it show time in (US-CANADA) time zoze . IT should be IST .

before selecting the date I have ti convert it firsrt into IST ?

indigo scarab
# indigo scarab Will look into that share results with you

regarding this I can't get past 57 wer I did exactly what was done in hf blog on whisper fine tune. Also I found out my dataset has 3-5s sample and some has noice. I also did spectro augment and other augment with +-2 pitch. Also, the CER decreases till 21 and wer reaches min 57 and then it explodes while val loss decreases.

topaz nexus
#

is kaggle getting too much traffic today

#

for the first time i am getting queued on a p100

dreamy cloak
#

🚀 Looking for a Partner to Join Our AI Platform!

Hey everyone! My team and I are building an amazing AI platform with the potential to transform how people work and innovate. We’re searching for a driven, talented partner to join us on this journey.

What we’re building:

Cutting-edge AI tools for real-world applications

Scalable, intelligent systems powered by the latest AI/ML technologies

A platform designed to make work smarter, faster, and more impactful

Who we’re looking for:

Strong English communication skills

Able to work in the US time zone

Ready for long-term collaboration and committed to tackling challenges

Passionate about AI, software engineering, or product innovation

Comfortable taking ownership and driving projects forward

If you’re excited about AI and want to be part of a small, high-energy team shaping the future, we’d love to hear from you!

💡 Let’s innovate together!
Send me a message if you’re ready to take on this challenge.

chrome rover
dreamy cloak
rich haloBOT
#
blarinkoid has been warned

Reason: Posted an invite

tall sphinx
#

I recently published an Egyptian Real Estate dataset, and I’ve built it from the perspective of a researcher who needs clean, ready-to-use data. I’ve never been someone who chases ranks for the sake of it; for me, the value of the work comes first.

However, it can be disappointing when you pour your heart into a project and it goes unnoticed. Kaggle is what gives me passion and a reason to open my laptop every day. I’m currently just one upvote away from becoming a Dataset Expert, and hitting that milestone would be an incredible payoff for the dedication I’ve put in.

Please have a look at my work, and if you find it useful for your projects or analysis, I’d appreciate your support!
https://www.kaggle.com/datasets/waddahali/real-estate-listings

errant glacier
#

Hi, what is the best server and channel to casually discuss with others about the NVIDIA Nemotron Model Reasoning Challenge?

unreal warren
turbid crest
vestal jasper
#

Hi everyone 👋

I just published two new Kaggle notebooks while learning Image Segmentation and Generative Models in Deep Learning. , This are Projects, so You know you can use them perfectly 😚

📌 Topic 1:
Medical Image Segmentation using U-Net
✔ Encoder-Decoder architecture
✔ U-Net implementation in PyTorch
✔ Medical X-ray lung segmentation
✔ Visualization of predicted masks
✔ Understanding pixel-level prediction
📊 Key Result:
Successful segmentation of lung regions from X-ray images.

Notebook link: https://www.kaggle.com/code/debanganghosh/medical-imaging-project-u-net-segmentatation

📌 Topic 2:
Building a Generative Model (GAN) from Scratch
✔ Generator vs Discriminator explanation
✔ GAN loss and adversarial training
✔ PyTorch implementation from scratch
✔ Generating handwritten digits from noise
✔ Visualization of generated samples
📊 Key Result:
DCGAN successfully generated realistic MNIST digits.

Notebook link:
https://www.kaggle.com/code/debanganghosh/build-generative-model-from-scratch-gan-dl

If you find them useful, an upvote on Kaggle would really help my learning journey 🙌

Happy to discuss improvements or suggestions!

merry light
#

test

true nymph
#

hi

grim thistle
#

Hello every one

rancid rampart
#

Just shipped my second Starbucks data science project on Kaggle — this time a recommendation engine + new product optimizer.

8 notebooks, 2 models (both Usability 10.0), 100K synthetic transactions constrained by real FRED/weather/menu data.

The fun part: designing a new Frappuccino using constrained optimization, and building a content-based recommender across 5 customer personas.

Check it out: https://www.kaggle.com/datasets/shiratoriseto/starbucks-recommendation-engine

First project was a 15-notebook Manhattan spatial analysis series if anyone's interested in that too.

inland wren
#

Hello!

opal ferry
#

Hello

high gust
#

hi

opal ferry
#

Hru

robust orbit
#

Hi

high gust
#

Hello everyone

sullen burrow
#

Hi

opal ferry
#

What's going on guys

bright patrol
#

yo!

opal ferry
#

Hru

red hawk
#

Hello :3

flint aspen
#

bro dead chat

#

what should i take up first

wary oyster
#

Hello

onyx urchin
#

How kaggle consider daily activity and mark blue in heat map?

#

👋

rich haloBOT
#
kogi0716 has been warned

Reason: Bad word usage

#
kogi0716 has been warned

Reason: Bad word usage

#
kogi0716 has been banned

Reason: Too many infractions

floral kettle
tender granite
#

Guys ,does anyone know data analytics or data science,I need some reference and I have some doubts,if anyone knows pls free to dm me , 🙏

neat garden
#

hey yo

unique blaze
#

Hello, everyone
I am Ohio, a student of the University of Lagos, Lagos, Nigeria, and a lover of data analytics, although still a beginner; I hope to learn a lot from y'all.

magic dove
#

Hello everyone,
is anyone enrolled in WiDS 2026? we can maybe discuss and work together

formal arch
#

HOI

turbid crest
stone condor
#

Hi

opal ferry
#

Hello

eager ivy
#

Hi

neat garden
#

hi

turbid crest
#

hi

dawn wharf
#

Hi

deep wasp
#

Hi

eager ivy
eager ivy
#

@chrome rover

opal ferry
eager ivy
opal ferry
eager ivy
primal olive
#

hiiiiiiiiiiiiiiiiiiiiii

wooden sleet
#

Heyaa!!!

vestal jasper
#

Hi everyone 👋

I just completed a 10-Day Intensive Deep Learning Challenge and published multiple Kaggle notebooks documenting my learning journey.

📌 Topic 1:
Word Embeddings + LSTM for Sentiment Analysis

📊 Key Result:
~84% accuracy on the IMDB sentiment dataset.

Notebook link:
https://www.kaggle.com/code/debanganghosh/deep-learning-nlp-build-word-embeddings-lstm

📌 Topic 2:
Time Series Forecasting with LSTM

📊 Key Result:
Model predicts Apple stock trend using historical data.

Notebook link:
https://www.kaggle.com/code/debanganghosh/deep-learning-for-finance-lstm-stock-prediction

📌 Topic 3:
Transformer NLP from Scratch

📊 Key Result:
Built a transformer model for news classification.

Notebook link:
https://www.kaggle.com/code/debanganghosh/transformer-nlp-from-scratch-text-classification

If you find them useful, an upvote on Kaggle would really help my learning journey 🙌

Happy to discuss suggestions or improvements!

opal ferry
#

Does anyone know the classical splines Theory?

turbid crest
alpine tide
#

Hey everyone — doing some independent research
on how active Kagglers manage their competition
workflow day to day.

Curious about: how you track competitions and
deadlines, how you analyse your own performance,
and whether you have ever used any external tool
or workaround to fill gaps in what Kaggle
shows you.

3 minutes, fully anonymous, no product pitch.

https://bit.ly/kaggle-workflow-study

DMs open if you want to share more — genuinely
reading every response.

robust orbit
#

Hey everyone,

I’m looking for someone who’s interested in learning machine learning through real projects and innovation.

Plan:
We’ll pick specific days to meet (online)
Do brainstorming sessions
Research ideas together
Decide on a project → build it → learn along the way
Also explore internship/job postings to understand current industry demands and align our learning

About me:
Intermediate in Python
Basic knowledge of ML libraries
Built 3 projects so far
Strong interest in math
Familiar with supervised learning and basics of neural networks
Sometimes we’ll also go deep into the math behind algorithms, so interest in math is a plus.

If you have a similar background and mindset, just DM me.
Let’s learn together and build something unpredictable 🔥

potent kindle
clever yarrow
#

Hi, I’m Deepak Sethi, Founder & CEO of Howdy Analytics. We’re building a next-gen Productivity Engine for digital creators....designed to turn ideas into high-impact content effortlessly. Currently in user trial, it’s already delivering top-notch results. Would love for you to try it and share feedback: https://hellobusiness.info

opal ferry
#

Cool

#

Link is unavailable bro

#

Due to update in the app

obtuse ridge
clever yarrow
#

Mate, it’s working. I’ve checked. Could you please do a hard refresh? It seems like a browser issue.

dire crown
#

Hi

winged echo
#

How do I access the Feedback channel?

manic lava
chrome rover
manic lava
chrome rover
#

Hmmmmmm .... can you DM me a screenshot? It looks right in my permissions and testing, but I must be missing something

manic lava
chrome rover
#

OK, it looks like this is working now. Let me know if you have issues folks!

gloomy meadow
#

the new logo looks odd. Might be that i'm just not used to it, but the dark background looks weird

spice trench
#

Hello

manic lava
chrome rover
manic lava
sinful whale
crystal verge
#

🛡️ Just shipped an open source tool that might be useful to anyone working with AI-generated Python or JSON.

AI Code Integrity Auditor scans for the failure patterns AI code generators specifically produce — the stuff that passes syntax checks but breaks at runtime. At this time is is specifically for Claude and ChatGPT generated code.

Detects things like:

  • Hallucinated variable and function references
  • Silent exception handlers that swallow errors
  • Placeholder logic left behind as complete code
  • Functions that claim to return values but don't
  • Success messages with no actual state change
  • JSON schemas with duplicate keys and null-heavy templates

Built in Python with Streamlit. Free and open source.

Would love feedback from anyone who uses AI-assisted coding in their Kaggle workflows.

🔗 https://github.com/bellatrix11176/AI-Code-Integrity-Auditor

stuck bison
#

Did they pause the airborne challenge?

#

I was about to win the $15000 prize.

rich haloBOT
#
panther1148 has been warned

Reason: Posted an invite

manic lava
turbid crest
turbid crest
proud basalt
turbid crest
#

🚨 New Dataset: Shopify (SHOP) Historical Stock Market Data (2015–2026)

The Problem It Solves:
• Inconsistency: Eliminates data discrepancies across platforms
• Messy Formats: No more raw strings or broken dates
• Unreliable Training: Avoids "garbage in, garbage out"
• Wasted Time: Saves hours of data cleaning

The Solution:
• Multi-Source Validation: MarketWatch, NASDAQ, Yahoo Finance
• Fully Cleaned: Standard dates, uniform prices, clean volume
• Consistent Structure: Proper OHLCV format
• ML-Ready: Works directly with Pandas, Sklearn, forecasting tools

🔗 Dataset Link:
https://www.kaggle.com/datasets/anadiskt/shopify-shop-stock-market-dataset-20152026

knotty totem
#

We are Looking for these externatal Mentor Position
Suggest if you have anyone
Experience level: 2-3 years

1 Zero-to-One SaaS Founders (The 0-1 journey)
2 D2C / Consumer Brand Builders (Product & Brand)
3 Startup CTOs / Product Builders (Tech & Scale)
4 B2B Sales & GTM Operators (Revenue & Strategy)
5 Growth & Distribution Experts (Marketing & Virality)
6 Startup Finance / CFOs (Fundraising & Runway)
7 Operator-first Angel Investors (The Investor Lens)
8 Founder Mindset Coaches (Leadership & Mental Resilience)

neat garden
#

hello

past pecan
#

Hello

opal ferry
#

Hello

manic lava
#

Hİ

turbid crest
#

hello

little kettle
stark badge
#

Should we be polite to Chatgpt?

opal ferry
#

That's called positive promoting

#

When you like a model’s response and reinforce it with positive follow-ups, the model tends to continue in the same style—this is often called positive prompting. It doesn’t actually learn or update itself, but uses your conversation as context, so your feedback guides future outputs.

#

So whenever you like the response just reply with anything positive.

white quest
ruby hound
fickle vine
nimble vine
#

hello guys, I'm Brazilian, my projects aim Public Management Data from my country

stark badge
stark badge
stark badge
#

Which is not bad, I think that’s pretty good

#

Thanks for your advice

opal ferry
cinder pine
#

Hi all. Im building a new community based platform and I would really love to get some feedback. it will really mean the world to me to hear voices from different parts of the globe. feel free to DM me.
my vision: "The system should not be controlled — it should be allowed to become."

gloomy meadow
#

how's this related to the server? jerry_stare

digital lily
opal ferry
muted trail
chrome rover
opal ferry
#

Guys mlflow or WB?

#

I know mlflow but gemini and chatgpt are forcing me to use WB lol they say it's better for experiments and visually also. Any tips ?

fickle vine
opal ferry
fickle vine
opal ferry
#

Yeah wandb is wb right??

#

I was using mlflow but needed lots of setup lol it's more for production focused setup I guess

#

If you are at my place which would you choose for ml or dl only ( I am not considering llms or agentic ai)?

fickle vine
opal ferry
#

Yes yes

fickle vine
opal ferry
#

Really helpful 😄

#

☺️*

opal ferry
burnt urchin
fickle vine
opal ferry
#

We cannot Even do nested runs in weights and bias 🥹

#

We can just group them from the dashboard that's so messy

#

UI is also laggy

#

But I admit that the boiler plate is so less in wb

fickle vine
leaden rock
#

Yo

#

Guys

little kettle
neat garden
#

yo

fallow zenith
#

yo

wheat shoal
#

yoo

chrome rover
quick brook
#

Someone should build a prediction model to predict the probability of a "hello chain" or "yo chain" occurring here within 24 hours 😂

neat garden
ivory roost
#

Helyo

stuck python
#

Is anybody else getting 404 not found for when trying to open kaggle?

#

I keep getting either that error or some screen with a red error message for the past hour and I gotta submit my final solution for a competition 😭

stuck python
#

Ok, weird, turns out everything works fine if I use hotspot, but if I use my WiFi on any device or even with a vpn it doesn’t work

drowsy dragon
#

Hii 👋 guys

nimble vine
#

hiiiiii

outer osprey
outer jay
#

Hi

fading ivy
#

Anyone alive Here...?

boreal creek
#

Can anyone give referral or any openings for freshers. Please consider 🙏

hard kraken
ember leaf
chrome rover
fading ivy
eager ivy
#

hi guys

floral anchor
#

Hi @everyone! 👋

We’re running a quick career survey at Women Who Do Data (W2D2) and would love to hear from you.

No matter your background, industry, or experience level, your input will help us better understand career journeys and build more valuable resources for our community.

💬 Take the survey: https://form.typeform.com/to/iY01FVhH

It’s short, anonymous, and super helpful—feel free to share with others too!

ember leaf
vale mantle
#

I built something that I wish I had when I started learning Git.

Most Git tutorials either:

  • Overcomplicate things
  • Or just give commands without context

So I created a complete Git + GitHub guide that actually teaches you through real-world scenarios.

What’s inside:
• Beginner → Advanced Git
• Scenario-based learning (like real dev work)
• Merge conflicts, PRs, branching, undo mistakes
• Advanced topics like reflog, bisect, cherry-pick

Perfect for students, developers, and open-source beginners.

If it helps you, consider starring it — it really motivates me.

https://github.com/Krish6115/GitxGithub

warm marten
#

Hello everyone, how many people know PyTorch. I need some testers to test a project I am working on.

forest fiber
vagrant pecan
forest fiber
forest fiber
vagrant pecan
rich haloBOT
#
1nfxri. has been warned

Reason: Posted an invite

warm marten
tribal vector
opal ferry
#

Hello guys

neat garden
#

hi guys🤗

patent elbow
#

hello

#

guys, I am looking for spring boot course. Is there any that you can suggest.

opal ferry
#

Any good book on game theory?

stoic basalt
#

cat

midnight rain
#

hi

terse orbit
terse orbit
terse orbit
midnight rain
west bone
#

I am struggling to find a messy dataset to practice data cleaning. But i am tired of find a perfect dataset.
Could u help me to get that?

bright cradle
#

Hi @everyone 👋 ,
I'm aspiring Data Analyst and have joined kaggle to sharpen my skills.
Happy to connect and learn with you guys.

umbral geode
#

hello

obsidian light
#

Hey! I'm a beginner, looking for 1–2 study partners at a similar level to learn together, share notebooks, and eventually team up for ranked competitions. I'm learning Python + ML from scratch and putting in 4–5 hours daily. DM me if interested!

opal ferry
#

Hello

wintry geyser
#

hello

neat garden
#

hello

brazen spindle
#

hi

sonic sparrow
#

Hi everyone

sonic sparrow
candid leaf
#

Anyone tried the Agent exam yet?

white furnace
#

Calling all environmentalists & conscious individuals! 🌍

If you care about things like sustainability, clean living, planting trees, healthy lifestyle, skincare 🧘🏻‍♀️, or even discussing global & social issues .. ♻️

We’re building a community GreenHearties where you’ll find like-minded people.

A space to connect, share ideas, and actually make an impact

💚 Join here:
https://chat.whatsapp.com/G0MDEtAlwrwCKDC3gJyp9o 💡🌱

idle inlet
#

hello I am Anubhav Mishra
Aspiring Data Analyst | | Great interest in data science && AI (Artificial Intelligence)

rich haloBOT
#
kou3095 has been warned

Reason: Bad word usage

candid leaf
tall sphinx
#

Hey everyone,

I’m not sure if you’ve been following the discussions over the past two weeks, but I recently completed a challenge called "14 Days, 14 Datasets." The challenge is now over, but it resulted in several high-quality datasets covering highly relevant topics.

The final topic is very personal to me: my home country, Sudan. As many of you may not know, Sudan has been experiencing conflict since the '90s, though it was previously concentrated in the Darfur region rather than the capital, Khartoum. Since 2019, Sudan has faced widespread demonstrations and government crackdowns that deeply affected Khartoum. Then, in 2023, a full-scale war broke out in the capital.

This conflict began as an attempt by the Rapid Support Forces (RSF) to seize authority from the National Army. Backed by the UAE which has funded the militia to gain control over Sudan’s gold resources—this war has cost civilians everything: their homes, their cars, their life savings, and their lives.

Because of this, I decided to curate a high-quality dataset to provide information on the reality of what is happening in my country.

Dataset Link: https://www.kaggle.com/datasets/waddahali/sudan-conflict-2023-2026

The dataset is fully documented, and the description provides extensive context. I hope you take a look, and please keep Sudan in your prayers.

Thank you all!

odd escarp
vapid fjord
#

hello guys. they fixed the problem and now i can see and text messages

onyx meteor
sturdy atlas
#

Hey does anyone know where to find good, FREE online courses for programming or data science

tight kiln
wild birch
chrome rover
rustic flint
#

Hiii

opal ferry
#

Hello

hot cobalt
#

hi guys does anyone knows any tools to send many emails without getting a spam ?

proven iron
#

Hi Kaggle...

thorn slate
#

Hi

terse orbit
#

hello everyone

odd terrace
#

HI I'm Shayan (WORLDS Youngest Kaggle Grandmaster) at age 13

#

muhammadshayan5839
checkout my kaggle id

opal ferry
#

Hi

dreamy scaffold
#

Gemma 4 Key Capabilities

Thinking – Built-in reasoning mode that lets the model think step-by-step before answering.

Long Context – Context windows of up to 128K tokens (E2B/E4B) and 256K tokens (26B A4B/31B).

Image Understanding – Object detection, Document/PDF parsing, screen and UI understanding, chart comprehension, OCR (including multilingual), handwriting recognition, and pointing. Images can be processed at variable aspect ratios and resolutions.

Video Understanding – Analyze video by processing sequences of frames.

Interleaved Multimodal Input – Freely mix text and images in any order within a single prompt.

Function Calling – Native support for structured tool use, enabling agentic workflows.

Coding – Code generation, completion, and correction.

Multilingual – Out-of-the-box support for 35+ languages, pre-trained on 140+ languages

Audio (E2B and E4B only) – Automatic speech recognition (ASR) and speech-to-translated-text across multiple languages.

https://blog.google/innovation-and-ai/technology/developers-tools/gemma-4/
https://developers.googleblog.com/en/bring-state-of-the-art-agentic-skills-to-the-edge-with-gemma-4/

dawn gate
#

hey guys

#

SyntaxError: Unexpected token '<', "<!doctype "... is not valid JSON

#

I am getting this whenever I go to the page for gemma-4-good-hackathon

daring remnant
#

the link seems to be working for me

#

try this one

karmic quarry
#

Hello

severe storm
#

looking for team mates for ARC prize 2026 any or all tracks. final decision will be decided after discuss. only interested person can contact.

pastel sedge
#

Hii I got 20 upvotes that are counting from medals but still can't get the silver medal on my notebook or dataset its been a day can anybody help?

terse orbit
pastel sedge
#

or what can I do about it?

terse orbit
#

just leave it and not open those notebook until you can get medal

pastel sedge
#

got it 👍

wooden sleet
solemn maple
#

The new classification between competitions and hackathons confuses me

chrome rover
solemn maple
cosmic sigil
pastel sedge
inland pagoda
#

Hello room

opal ferry
#

Hello guys

bronze sapphire
#

Hey everyone 👋

I’m building something meaningful in healthcare and would genuinely love your support.

We’re working on OraCAN — a portable AI-based device that helps in early detection of oral cancer (even in remote villages). The goal is simple: detect cancer early and save lives.

We’ve just started growing our presence on LinkedIn, and every single follow helps us reach more doctors, researchers, and people who can benefit from this.

If this sounds worth supporting, it would mean a lot if you could follow our page 🙏

👉 https://www.linkedin.com/company/unmolya-medvision

No spam, just real progress and impact updates. Appreciate you all ❤️

thorny robin
#

Hello

rich haloBOT
#
chico6069 has been warned

Reason: Posted an invite

tidal nymph
#

how hard is it to win a competition

hazy spruce
#

Hey everyone, i'm new

rocky hare
#

anyone from ai developer???

#

anyone interseted to build with ai project with me??? im already working on it i need backend developer as well fronted litter???

plush rain
#

Wow

#

I got verified

frosty gyro
serene quail
#

hello everybody

severe storm
#

Hello, looking for AGI ARC prize competition, team mates

cold pilot
#

Need your honest take, 60 seconds

Hey everyone 👋
need your real opinion, if you've ever used AI pls.

https://forms.gle/GynGGBWF46de9jz47

4 questions. Would be supper happy for every next sumbit!
This is non-commercial. Only personal interest.

frank shore
#

hi, i have started the gemma4

severe storm
proven iron
#

Hello kaggle.

solar wind
#

Happy Weekend!

Hello Everyone!
If you know someone who have good skills in Python and Machine Learning, Please invite me!

Our Company is open to hire Python and Software Engineer.

Requirements:

  • 2+ years of Software Engineering Experience
  • C1 or Native English Level
  • Good vision of Software Trent

Benefits:

  • Competitive Income
  • Supporting Several roles and chances
  • Multiple Role Working is enable

Important:

  • Our company is designed for Capability Person.

Questions:

  • For Junior Persons?
    Do not give up, strong enthusiasm is also big point and our company also focus on the person's enthusiasm.

Thanks again.
Sophia

vivid locust
#

Hi Kaggler!
This is Mary from US.
I like exchange knowledge with Intelligence Persons.

Lovely World!

thorn slate
potent kindle
rich haloBOT
#
its_nilaysaha has been warned

Reason: Posted an invite

uncut schooner
#

Hi

frozen hatch
#

Hello, everyone anyone from Bangladesh and instead in project pin me

#

Cash prize 🏆

marsh carbon
#

@everyone
Namaste!
Rebel from INDIA<3

opal ferry
#

Hello beautiful people

solar wind
#

Hello Everyone!
If you know someone who have good skills in Python and Machine Learning, Please invite me!

Our Company is open to hire Python and Software Engineer.

Requirements:
2+ years of Software Engineering Experience
C1 or Native English Level
Good vision of Software Trent

Benefits:
Competitive Income
Supporting Several roles and chances
Multiple Role Working is enable

Important:
Our company is designed for Capability Person.

Questions:
For Junior Persons?
Do not give up, strong enthusiasm is also big point and our company also focus on the person's enthusiasm.

Thanks again.
Sophia

chrome rover
deft kindle
#

Guys support my youtube channel

mental roost
#

Hello

heady flax
#

Hiii

bleak crater
#

Hi everyone!
I’m Titim , a beginner in Machine Learning and Python. I usually work on projects on my own, but I thought it would be fun to team up with others and learn together while building beginner-friendly projects.
The goal is to share ideas, experiment, and enjoy the process—projects could be small ML models, data visualization, or anything beginner-friendly.
If you’re interested in collaborating and having fun while learning, reply here! Thanks in advance, and I’m excited to meet new teammates!

chrome rover
#

Show us what you're proud of!

opal ferry
#

Hello guys

faint cipher
slender dock
#

hello？

hazy spruce
high jewel
#

Hii

sleek nexus
#

hii

solar wind
#

Hello Everyone!
If you know someone who have good skills in Python and Machine Learning, Please invite me!

Our Company is open to hire Python and Software Engineer.

Requirements:
2+ years of Software Engineering Experience
C1 or Native English Level
Good vision of Software Trent

Benefits:
Competitive Income
Supporting Several roles and chances
Multiple Role Working is enable

Important:
Our company is designed for Capability Person.

Questions:
For Junior Persons?
Do not give up, strong enthusiasm is also big point and our company also focus on the person's enthusiasm.

Thanks again.
Sophia

barren jolt
#

hi

#

anyone has connection with her @restive ice

#

very emergency !!!!!!!!!!!!

verbal rune
#

hello everyone

#

She can't read. She can't type. She has no internet.
We put an Android phone in her hands. Airplane Mode.
She spoke Luganda. The AI responded. Completely offline.
Then she said 5 words that changed everything for us.
This is our founding story.
Watch from 2:16:
🎬 https://youtu.be/cs9PGfO1eo0 🔗
Personal LinkedIn: https://www.linkedin.com/posts/katostevenmubiru_gemma4-tinyaya-sovereignai-activity-7448003053439610880-tawo?utm_source=share&utm_medium=member_desktop&rcm=ACoAABtmWdcBpUgQ-vJgDJ__uOH7hIZtb_436VA
Company Linkedin: https://www.linkedin.com/posts/crane-ai-labs_where-the-cloud-cant-reach-crane-ai-labs-activity-7448003034372186112-6muf?utm_source=share&utm_medium=member_desktop&rcm=ACoAABtmWdcBpUgQ-vJgDJ__uOH7hIZtb_436VA

X: https://x.com/katomubirusteve/status/2042237343432835362?s=20

Crane AI Labs 🇺🇬 Where the Cloud Can't Reach. We Can.

spiral nova
rich haloBOT
#
codex0103 has been warned

Reason: Posted an invite

#
gravy6819 has been warned

Reason: Posted an invite

rich haloBOT
#
radyo_activ has been warned

Reason: Posted an invite

#
macny0114 has been warned

Reason: Posted an invite

dreamy cloak
#

hi

unreal storm
#

Hi I m AI/ML Engineer and new to this server!

silk nacelle
#

Hi

#

I am new here

rich haloBOT
#
playful_quokka_37450 has been warned

Reason: Posted an invite

#
lupegizmo_02338 has been warned

Reason: Posted an invite

#
pratiksha_7777_74568 has been warned

Reason: Posted an invite

#
saran85445 has been warned

Reason: Posted an invite

fluid crystal
#

hi

#

is there any way to get on-site support? @restive frigate

glad basalt
#

check my bio 😁

rich haloBOT
#
iron.k has been warned

Reason: Posted an invite

acoustic frigate
#

hi i am bengazy-dev learning to code

rich haloBOT
#
0x00syfer has been warned

Reason: Posted an invite

versed raven
#

Hey guys ! I've got Coursera plus and edX premuim at cheaper rates . 2000 rupees for 1 year. Dm to know more .

opal ferry
#

Hello

unreal storm
#

Hi

outer quartz
#

Yooo

#

My model just jumed from loss 2.4 to 1.9 and then 1.6

#

[nanothought]: Step 14950 | Phase: PRETRAIN | Loss 2.4727 | LR 8.31e-05 | Cycles 1
[nanothought]: Step 15000 | Phase: PRETRAIN | Loss 2.4335 | LR 8.30e-05 | Cycles 1
[nanothought]: Step 15050 | Phase: PRETRAIN | Loss 1.9250 | LR 8.28e-05 | Cycles 1
[nanothought]: Step 15100 | Phase: PRETRAIN | Loss 1.6684 | LR 8.27e-05 | Cycles 1

#

Im pretraining it on FineWeb-Edu

#

Basically I just want to create a better version of nanochat

#

and its 1 bit

lyric cedar
woeful slate
#

Hi

pliant kayak
#

hii

amber vigil
#

Hi guys, I just made a server for entirely quantitative trading competitions. DM for access since I don't think I can post an invite here

unreal storm
#

Hi

ornate snow
#

Hi everyone, I’m Samia Ghulam. I’m an aspiring Data Analyst with hands-on experience in SQL, Tableau, Python, and R. I like working on data cleaning, analysis, and dashboard projects, especially around sales, inventory, and business insights. I’m here to learn, share, and grow with the community

unreal storm
#

Hi guys rate my website from my profile, any improvement, recommendation etc?

unreal storm
#

Hi

digital oasis
#

Hello

half relic
#

Hi I'm Rishi, eager to learn and grow

civic granite
#

Good day, I am a student working on his project which deals with fruad detection in fintech

lofty pike
#

Hello. I am noob. Only know p-hacking.

rich haloBOT
#
eric900 has been warned

Reason: Posted an invite

quick brook
rich haloBOT
#
anandh6487 has been warned

Reason: Posted an invite

neat garden
#

hi

left ermine
#

hey guys

#

so im tryna build a blackboxing sovereignty layer, i wanna connect with a few people who know how ai drifts with terms, im trying to understand how to prevent hallucination and ip leak by these blackboxed terms without full on abstraction that is so abstract that a frontier cloud model cannot reason properly

rich haloBOT
#
avdhesh_varshney has been warned

Reason: Posted an invite

unreal storm
#

Hi guyz

keen hull
tight zealot
heady flax
#

Hiii

#

Iam frk

outer quartz
#

Hello! Could someone please explain why is the tpu accelerator option blanked out

kind kelp
#

Helllooooo 👋

midnight rain
#

hi!!

unreal storm
#

Hi

peak nymph
#

This is an Ai- Jarvis...which can talk in any language
Waana try !!

static axle
#

Hi everyone

unreal storm
#

Hi

honest basin
#

hi

compact lotus
#

Hello
I am participating in a hackathon and I need some help
It is about an insurance company that wants to analyze how risky it is to assure a place depending on the the seismic classification it is
I have data that I don’t understand very much and I am supposed to create a dashboard showcasing aces where they might lose , win.. (GIS)

#

If you have come across any similiar project please tell me

solar wind
#

Hello Everyone!
If you know someone who have good skills in Python and Machine Learning, Please invite me!

Our Company is open to hire Python and Software Engineer.

Requirements:
2+ years of Software Engineering Experience
C1 or Native English Level
Good vision of Software Trent

Benefits:
Competitive Income
Supporting Several roles and chances
Multiple Role Working is enable

Important:
Our company is designed for Capability Person.

Questions:
For Junior Persons?
Do not give up, strong enthusiasm is also big point and our company also focus on the person's enthusiasm.

How to apply?
DM with resume and 1min's record of your English Speaking

Thanks again.
Sophia

winter mountain
#

Salamoona

chrome rover
flint ledge
#

hello everyone

#

how are you all

unreal storm
#

Hi everyone, I’m an AI/ML engineer passionate about helping others and collaborating on open-source projects. If you share similar interests, feel free to connect.

torn oriole
rich haloBOT
#
pritam351645 has been warned

Reason: Posted an invite

rich haloBOT
#
travis_s2005 has been warned

Reason: Posted an invite

quick brook
#

Is there any such framework where multiple derivatives are automatically taken, multiple integrations occur to try and maximise term diversity, to which multiple occurrences of gradient descent on the coefficients can occur (one instance per level, where say you find two different sets from a derivative, that is two different levels) to enable automatic refinement of derived formulas? Seems like a longstanding problem and an obvious solution, so I'd presume something already exists? Persisting derivatives can be flagged for possible cyclical behaviour and resolution of these cycles could automatically be sought as well

boreal cargo
#

hello we are acestrial group of developers and we creted our new computer language and first phase is live now

rich haloBOT
#
ayushjha0373 has been warned

Reason: Bad word usage

boreal cargo
#

HumanSpeak v1.1.2 is live

HumanSpeak is a native, readable programming language created by Ayush. It is designed to feel simple like English, but still be useful for real software such as apps, websites, automation, files, web servers, and AI-style workflows.

Download and learn it here:

GitHub repo: https://github.com/ayushjhaj8-sys/HumanSpeak

No Python needed at runtime
Simple English-like syntax
Works on Windows, macOS, and Linux
Includes an installer for each platform
Supports tasks, classes, files, web apps, errors, and project commands
I

halcyon gale
#

yo everyone

#

wsp

opal ferry
#

💀💀 is that a new algorithm

frozen hatch
#

Hey!

wise seal
#

hello friend

silk nacelle
#

Hi Everyone, any news about upcoming Feture competition date 📅

crimson crystal
#

d

#

Hello, can i have a Txt To Speech from HuggingSpaces that works perfectly? Cuz Huggingspaces models is having tons of errors

If anyone can help me please mention it because i need Txt to speech with voice cloning

(Im not spamming this lol)

crimson crystal
umbral patrol
#

hey there i am new to this !

dense wedge
#

Hey there, so do i !

near belfry
#

:D

dire drift
short pumice
#

well...

chrome rover
ripe ridge
inland grove
#

Will it be like a mix of coding and vibecoding or just completely giving in to the vibes?

#

I see, it's completely vibes

hasty stump
rich haloBOT
#
hetesh_ has been warned

Reason: Posted an invite

keen hare
strong dawn
#

Why was the announcement deleted ?

keen hare
#

i am guessing because most of the reactions were 💀

chrome rover
#

LOL - no. There's a minor issue we're fixing.

opal ferry
#

Hello guys

umbral quartz
keen hare
rustic flint
#

Hello guys sup

chrome rover
inner scaffold
#

ey yo?

rich haloBOT
#
rianrafid has been warned

Reason: Posted an invite

obtuse plume
#

Heyo

ember leaf
#

Hiya guys

jovial pollen
#

howdy

digital ledge
warm pasture
#

I Signed up for it!

unreal warren
glad echo
#

hello guys,
I'm have been learning ML since one year, I want to be an AI engineer . I am not from engineering college, i'm learning alone, So that there are many mistakes in my learning. And also there are too many distraction to work so I couldn't give time properly, .please give me some suggestion that will help to become an AI engineer.

unreal oracle
#

Hey guys, i need machine learning best playlist or video pls recommend me

past aurora
tardy dirge
#

Hlo

solemn bane
#

hello

dry garden
#

Hi Kaggle Team , I need your support, when i am trying to verify my account using my mobile number , it says that i am using multiple accounts. Can you guys please help me know what that other account is so that i delete it , or can you please delink my mobile number from it ?

chrome rover
keen hull
#

Hello everyone would love to show a new project where I built an MCP server for research papers that gives LLMs actual superpowers for academic research — here's what it does:

  • queries Semantic Scholar + OpenAlex APIs for real paper data
  • has Google Search (SERP) so it can search Google
  • can read full webpages via JINA reader
  • hosted on Hugging Face Spaces — just plug the link into Claude Desktop, no setup needed

Basically, your AI stops hallucinating and starts pulling real data from the internet

MCP Server: https://github.com/Sauravroy34/ResearchPaperMCP

Also built a React chatbot powered by Gemini that uses this MCP server under the hood:

  • built to improve both accuracy and interpretability of research answers

Research Agent: https://github.com/Sauravroy34/ResearchAgent

would love feedback / thoughts on where to take it next

small leaf
#

Hello all, I am data scientist working in the industry for 9 years now, I have built a branching notebook interface and runtime in Rust which allows you and your agents (claude/copilot etc) to work on data science tasks together! It is open source 🙂 would love your feedback/contributions https://www.tine.sh

rich haloBOT
#
adix0759 has been warned

Reason: Bad word usage

winged kite
#

Built in 30 Days. Now it’s coming for you.

We’re just 3 indie devs. No studio, no team—just us.
In 1 month, we built a full psychological horror experience in Unreal Engine.

ASTITVA: LOST WITHIN
30–45 minutes of guilt, distortion, and something that doesn’t let you leave.

Now it’s in post-production.

This isn’t just a jumpscare horror.
This is the kind that stays in your head after you quit.

We need your eyes on this.

What do you think of the vibe?
Does this feel different… or just another horror game?

—

Follow us everywhere.
Subscribe on YouTube.
Follow on Instagram & Threads.
Join our Telegram & Discord.

This is just the beginning.
Stay tuned!!!
#AntarStudios #AstitvaLostWithin #indiegames #gamingcommunity #psychologicalhorror

rich haloBOT
#
pochikon07 has been warned

Reason: Posted an invite

winged kite
primal nacelle
#

Test

charred ice
#

hi

forest lake
#

What is the duration of the capstone project?
Capstone Project Goes Public and Live: End of Day, Friday, June 15th, 2026
Submission Deadline: Sunday, June 30, 2026, at 11:59 PM PT

re: 5-Day AI Agents: Intensive Vibe Coding Course With Google
Course duration: June 15 - 19, 2026
I think the Capstone start date is a typo.
Clarification please: Jun 15th is Mon. Fri is 19th.

chrome rover
chrome rover
boreal umbra
small leaf
white furnace
#

Women often stay away from real-world knowledge, not by choice, but because access, exposure, and space haven’t always been there.

We’re working on changing that.
Building a community where women learn, share, and grow in areas like:
Geopolitics, Politics, Engineering, Finance, Investing, Cars and other spaces we’ve been kept away from. 💹

This is about becoming more aware, more independent, and more capable in real life.

Topic-wise groups will be created
Suggestions are welcome. Join:
https://chat.whatsapp.com/DIMuHRkriIMJ3NLTAmEBaf

rich haloBOT
#
alhaitham_69 has been warned

Reason: Posted an invite

lime epoch
#

Hi, nice to meet you all!
i want to be a prompt engineer

wise seal
#

can someone vote my notebook

opal ferry
#

Hello

ripe badge
#

hello everyone, i am Utkarsh, i am 16 years old and in my junior year of high school. I aspire to be an engineer, and I love computers and machines. Hope ill get along with everyone!

lime crow
#

Hello everyone

rich haloBOT
#
maitrak_manu has been warned

Reason: Posted an invite

verbal folio
#

I am trying to add my teamates to a competition but when searching it's not showing their names. Does anyone what to do?

gleaming bone
#

looking for an software engineering/AI/ML internship any leads??
skillset:
Languages: Python, JavaScript, TypeScript, Java, C/C++, SQL
Frameworks & Libraries: React, Next.js, Node.js, FastAPI, PyTorch, Pandas, NumPy
Databases & ORMs: PostgreSQL, Prisma
Tools & Technologies: Git, GitHub, Tailwind CSS, Vercel, Modal, Gradio
Core Areas: Full-Stack Development, Machine Learning, API Development, Real-time Applications, Model Deployment

pale dragon
exotic viper
#

is there a way to get more gpu compute on kaggle?

#

the 30 hours on kaggle is not enough

exotic viper
dawn prism
#

hello

#

mb i guess i posted this in wrong channel

upbeat finch
winged lion
#

Where I can share images in this server ? 🫪

lilac birch
#

Hi

#

Why are there no memes?

winged lion
#

I m trying to train a model but after half an hour kaggle starts to like freezing my instance i tried 2-3 times happening with me again and again

rich haloBOT
#
sujal__31 has been warned

Reason: Posted an invite

rich haloBOT
#
darklynx_3019_95561 has been warned

Reason: Posted an invite

night shoal
#

Hello everyone
in a competition I tried solving it normally and got a public score of 0.85. But then I searched for the dataset used in the competition and managed to find it. Using that dataset, I generated a submission file that scored 1.00.

Now my question is:

Is it possible to recreate the submission file using only the provided train and test datasets, without relying on the external dataset I found?

In other words, I want to understand if there is a way to learn or reverse-engineer how to produce the same submission output (ID → label mapping) using only the original train/test files. I’m not sure if “reverse engineering the submission” is the correct term, but I want to figure out how to get the same result properly using machine learning rather than external data.

Also, I want to clarify that for the submission I made, I actually had access to the full feature set—not just IDs and labels, meaning the other feature of the sub file

I would really appreciate any help or guidance. If needed, I can share the train/test files or the submission file that achieved the 1.00 score.

Thanks in advance!

lilac birch
#

Hi

#

Any kaggler online?

#

How much is one kaggler making every month?

#

$100,000?

#

Does a kaggler get more than 3 babes every month?

#

What is the color of your Bugatti?

#

11k online and noone is sharing color of their Bugatti?

chrome rover
full nest
#

Hi, I’m curious about the production workflow behind building AI agents in the course 5-Day AI Agents. Beyond getting an agent working in a demo, what would a real production workflow look like? For example, how should we think about testing, prompt/version management, tool/API reliability, monitoring, retries, guardrails, and deployment when moving an AI agent from prototype to production?

chrome rover
thorn slate
#

I am looking for LInkedin account manager for our team.

green bough
#

Hi All

lilac birch
hollow folio
#

Hey guys

I am a second year student who wants to learn the Ai ml and data science

Can u guys pls suggest me the free resources for learning this topics

inner sinew
#

hello

proud dragon
#

Hi everyone joining Kaggle channel on discord

light helm
#

Hi all! Haley here. Based in Melbourne AU. Have been working in financial service for 8 years and have been embracing the AI wave and building agents recently. If you are in AU or APAC let's connect.

sonic sparrow
unreal storm
#

Hi Guys I m AI/ML Engineer

chrome rover
solemn iron
#

Hello guys I am looking for a job and I am not able to find one and heaven knows how miserable i am 😭🙏
Anyone if u have backend intern opening orr golang intern opening
Refer mee so god can blast you with an insane package

toxic yacht
#

hi

tame crypt
#

👋👋

spring lotus
#

Hello guys I'm new here. Hoping to get the most from being here.

tired crown
#

hello! I joined to go the google 5 day thing too.

chilly swift
#

Check my bio 😁

tough coral
#

Hi

gray field
#

Hi everyone！I am new here 😀

pliant cape
#

hi

craggy comet
#

hi everyone!

minor wasp
#

hi

chilly swift
#

Yo guys my account was hacked ignore any link sent through this account

hard kraken
gentle narwhal
#

Hey

tropic stream
#

HELLOOO

blazing dragon
#

Hi everyone I am new and ready to learn 🙂

urban surge
#

Same! Here and eager for the knowledge! all_the_things

lyric pollen
#

hi everyone i am just a kid who currently mad a 50k project and because cybersecurity tools require trusted background now it i sitting in a private repo

#

mad/

#

made/

unborn siren
#

HI everyone i am new and ready to learn

finite tinsel
unique wind
steady pendant
#

Thanks

silk timber
#

Good morning, everyone. I'm glad to be here

magic pendant
#

Wassup

#

U guys sharing LinkedIn here

#

Dayum

dapper garnet
sour holly
#

Hey everyone!👋

I'm helping source respondents for an academic ML research project on student performance prediction. Looking for current university/college students to fill out a short survey.

✅ 29 multiple choice questions
✅ 2–3 minutes
✅ Anonymous
✅ Legit academic research

https://docs.google.com/forms/d/e/1FAIpQLSecrq6yt4J72NmctMDq9_Tt7YGYRifl2wOqN5QwWDlApbleIg/viewform?usp=pp_url&entry.1714577203=agent1

Would really appreciate it if you could fill it out and share with any student friends! 🙌

tropic stream
#

Hello

tropic stream
#

Im planning to make an AI R&D organization (non profit as of now) to focus on efficiency and accuracy of models below 50B parameters

#

focusing on mathamatical accuracy and real data

#

Any idea where I can find people who are interested to volunteer and how to scale

rocky glacier
#

Hi everyone! i am new and ready to learn

left cave
#

Hi, I'm anton. I'd love to build any automation pipeline

silver yew
#

Hello, I'm a passionate B.Sc. Data Science student with a keen interest in connecting with professionals and enthusiasts in the field. I enjoy exploring diverse aspects of data science and strive to gain knowledge and hands-on experience that can propel me forward in my career. Collaborative learning and sharing ideas with like-minded individuals inspire me, and I look forward to building strong networks while contributing to innovative projects. Let's discuss and discover the power of data together!

tribal tinsel
#

Hey everyone! Nipun here from India 🇮🇳
Background in Trust & Safety + community management, now shifting into AI system design. I’m particularly interested in building agent-driven workflows (RAG, automation, etc.). Here to learn fast, build fast, and connect with people doing the same 🚀

muted marsh
#

Pls show some love every one👋

thorn vale
#

Hey everyone! I am Shashank, ex-Amazon, IIT Guwahati alumni, and a vibe coder who has shipped multiple AI products. Currently exploring the intersection of AI and biology, drug discovery, and personalized medicine.

Building Project Rosie for Gemma 4 Good hackathon: https://www.shashankpadala.com/blogs/project-rosie-cancer-vaccine . A man used ChatGPT and AlphaFold to design a cancer vaccine for his dying dog and it worked. I am trying to build the infrastructure to make that repeatable.

Looking for collaborators with background in computational biology, bioinformatics, vet oncology, mRNA synthesis, or someone who can animate the science the way cartoon show "Ozzy and Drix" explained the human body as a kid. Wanna use that as inspiration for creating project demo video.

Happy to be here, excited to see what everyone is building!

finite tinsel
fluid crystal
#

anyone working on Gemma 4 Good?

ember halo
#

Hi everyone! I’m Alfina, a software developer currently exploring data science and cloud technologies . I enjoy learning new things and building my skills step by step. Lately, I’ve been diving into AI concepts and improving my problem-solving through platforms like Kaggle.

high skiff
#

Hello everyone, just joined and am starting in the field. Hoping to get the most out of this. Economics graduate here!

winged pasture
#

HI everyone ! I'm Attika, A CA student with different courses in AI. It's something of my interest so I want to learn more through kaggle. Wish you all a happy journey

verbal trail
little kettle
ember lantern
#

😃

winter kite
#

hii there

amber oxide
#

Hi all, I am interested in AI and want to land a job in the same this year. Would love to do projects with you all!

white wasp
#

Hi Everyone!

rough acorn
#

Heyy everyone

opal ferry
#

Guys can anyone tell me how to write solution writeup for kaggle competition?

south ocean
#

Hello

rich haloBOT
#
peaceloving has been warned

Reason: Posted an invite

royal jewel
#

Heyyyy

pulsar glade
#

Heyyy had a question.. the Kaggle and Google Agentic AI course - is that completely free?

#

or is there some "hidden" payment?

silk venture
#

Hello everyone, Ram here. Excited to learn more about data science and cloud. Thanks!

fallow otter
#

Hello everyone! New hobbyist here! Hope to learn more from the community here!

grizzled sentinel
opal ferry
amber turret
#

Hello everyone! My name is Forhad, and I’m from Bangladesh. I work as a Software Developer and am currently exploring data science and AI to expand my knowledge. I’m still new to Kaggle, but excited to learn and grow in this space.

fallow wedge
cyan ingot
#

Hey everyone! Rudresh here.
I am a student who has not yet obtained a bachelor's degree. I am here because I want to figure out what I should do next. Since many people here hold excellent positions in the fields of tech and AI, I would like to seek your advice on what subject I should pursue for my studies. I am from India 🇮🇳.

hollow pebble
#

Hii

rare grotto
#

Hey folks,

I am here to connect boroader community and stay on top of the IT filed. Thanks

bitter wasp
#

Hello guys. Hope you're keeping in a good shape. I'm a beginner in data scientist learning.

dawn prism
#

Hello everyone! New hobbyist here! Hope to learn more from the community here!

wild island
#

Hey @everyone, use this dataset for the new EDA + model build:

https://www.kaggle.com/datasets/vedantbhavsar43/ipl-2007-to-2026-complete-ball-by-ball-dataset

This is a better base than the older IPL datasets because it already has:

  • latest available IPL 2026 data
  • full ball-by-ball coverage
  • cleaner ML-ready structure
  • better feature engineering scope

The main advantage is that we can skip a lot of cleaning and directly focus on:

  • EDA
  • feature engineering
  • stronger model building

It should be much better for winner prediction, score prediction, and live match modeling.

little nova
#

Hello guys I am new here

charred dew
#

Hello all 🙂 Just an old man learning new tricks - Been testing how to use local LLM models for coding and automations, stumbled across Kaggle and decided to join to learn more about data/machine learning!

rich haloBOT
#
cherry0102992 has been warned

Reason: Posted an invite

rich haloBOT
#
orgiloogii0529 has been warned

Reason: Posted an invite

quartz oar
#

Hello Folks....,
I'm here.!

half basalt
#

Hello, I have joined today.. very new to Discord too

normal vapor
#

Hello everyone, I am new here. Any tips or suggestions?

carmine herald
#

hello everyone i am new here i want to learn ml and deep learning any suggestions or tips

cerulean dome
#

Hi everybody! Musa here

lilac birch
#

Hello everyone does anyone wants to teamup with hackathons here. I am an experienced data scientist working in bank.

#

@everyone

dense summit
#

hey

wary roost
#

Hi All,
I am pretty new to discord as well as data science.
I have worked as a Full Stack developer for a year and looking for a role.

Would appreciate any advice in making the switch.

ripe sleet
tropic stream
#

hey

eternal grotto
#

Hi

#

Maryam from Abbottabad, Pakistan

#

I'm a student of final year of graduation 🎓

dry finch
#

test

sterile shore
#

..

calm mason
#

Hello everyone ! My name is Vaishnavi from India, glad to introduce myself to you all i am a 2nd year student looking forward to explore and study , excited to learn something new and grow with you all, hope to learn so much from you all.

verbal trail
#

Hi @all , Mayank Sharma this side , 2nd year student at IIT Jodhpur . I am building a SaaS product named "SentinelNexus" an Automated SOC . Nice to meet you all looking forward to interact with great minds ! '

oblique isle
#

Test

plain roost
#

So many new people

#

Oh new course

#

that's why mhmm

bronze robin
thorn slate
#

📢 Hello Everyone!

I’ve been searching for someone with strong communication 💬 and development skills 💻 to join our team for the past month. Unfortunately, I haven’t yet found the right fit 😔. Many candidates applied, but their skills or experience didn’t match our team’s needs.

If you know someone who would be a good fit 🤝, please feel free to DM me 📩. Thank you for your help! 🙏

errant temple
#

Hi everyone,
​I’m an independent researcher and I’ve just finished a paper on Assumption-Gated BioDesign Runtime:
A Provenance-Aware Architecture for Safer Cross-Tool Reasoning
in Synthetic Biology Workbenches.

​I’m planning to submit it to arXiv under the cs.AI category, but I currently need an endorsement. I was wondering if anyone here with the necessary permissions would be willing to help me out?

​You can check out my work/preprint here: https://github.com/zhangze1007/Assumption-Gated-Biodesign-Runtime-Artifact

​Endorsement Code: OEOMPS

​Thanks in advance for your support!

lean plover
#

Hey folks! Jack here. I'm a CS & Economics student with a passion for building tools and digging into data. My core stack includes Python, C++, and Java, and I've spent a lot of time recently building custom desktop utilities and exploring AI Agent frameworks (like MCP). I'm joining the server to shift some of my focus toward predictive modeling and quantitative analysis. Looking forward to discussing data science, sharing ideas, and maybe teaming up for future competitions

sonic sparrow
cyan ingot
worn trellis
#

Hello guys!! Robert here from Ghana. I'm new and hope to learn from you all

strong wagon
#

hii

sudden maple
#

#👤┊introductions Hi all, this is Irfan. I am a QA Engineer working in one of the MNCs. I have joined Kaggle recently and look forward to meeting and engaging with you all. Let's have a good chat and move ahead with the learning.

glossy badge
#

@sudden maple
Hello...
As a fresher looking to join an MNC what are the must have skills beyond just coding??? I'd love to hear from those already working in the industry....

grim estuary
#

Hello everyone!

wicked sable
#

Hello everyone
I help businesses build fast, reliable web applications and AI-powered tools that automate work and improve customer experience.
I’ve experienced such as responsive web apps that work smoothly across all devices, secure systems that handle large amounts of data, and platforms that connect seamlessly with payment services and third-party tools. I also build AI features like chatbots that answer customer questions, dashboards that turn business data into clear insights, and automation systems that save hours of manual work every week.

robust bone
#

guys please any body help me, i have the dataset i need to build the model for it but what ever i try it never imrpoves it plateaued at 85.1

barren pivot
#

Hi

woven granite
#

Hello!

zenith ermine
#

hi

gleaming bloom
#

Hello everyone, I am new here willing to get deep in ML, Deep learning, Multiagents deployment.

tropic stream
#

Hey everybody

olive kite
#

Heyy 👋

arctic sonnet
#

Hi everyone,

I’m currently competing in a data science challenge and aiming to improve my leaderboard performance. I’m a self-taught beginner, but I’m fully committed and actively working to get better results.

I’m looking for someone experienced who can guide me, review my approach, or point out what I might be missing to improve my score. Any help, feedback, or direction would be greatly appreciated.

Thank you!

elfin wasp
#

Lemme show there is someone here 🙋🏻

elfin wasp
# robust bone guys please any body help me, i have the dataset i need to build the model for ...

Hey bro, if your model is stuck at around 85% it usually means you’ve hit a common plateau where the model already learned the basic patterns, so now you need to improve either the data or the approach—first check if it’s underfitting or overfitting by comparing train vs validation accuracy, then try improving data quality (clean labels, balance classes, add more data if possible), and after that experiment with better models like Random Forest or XGBoost and tune hyperparameters like learning rate, batch size, and epochs instead of using defaults; also focus on feature engineering and error analysis because most improvements beyond this point come from understanding where the model is failing rather than just changing the model blindly.

plain roost
vital root
#

Hi guys , i didnt train my model but instead i used it in a mobile app, and i created embeddings which my model use as reference is its answers, is that euqally good as the training ? in terms of comptetition

bright oriole
#

Hello Everyone, I am Francis Obama a petroleum engineer aspiring to be a data scientist or ML engineer in the oil and gas sector. I aim to learn and collobarate with people in the community. I am thinking to participate in the ROGII WELLBORE GEOLOGY PREDICTION Competition, still looking for teamates as it is my first time to participate in the ML or datascience competion.

elfin wasp
# vital root Hi guys , i didnt train my model but instead i used it in a mobile app, and i cr...

What you did is not actual model training. You used a pre-trained model and added embeddings as a reference system, which is more like a retrieval (RAG) approach. In real training, the model learns and updates its internal weights using data, but in your case the model itself doesn’t change, it just fetches relevant information from embeddings to answer. This method is still very useful in real apps, but for competitions it depends on rules—if they ask for training or fine-tuning, then embeddings alone may not be considered full training.

sterile jolt
#

Hello

upbeat finch
dense pivot
#

Has anyone automated the daily life tasks using Python?

vestal moon
#

hii

vital root
elfin wasp
vital root
#

thank you

thorn slate
#

Hey everyone! 👋
We’re looking for someone with strong communication skills and a solid educational background.
Experience we love: Hands-on experience with the client interview process.
Location: Fully remote 🌍

If you’re interested, shoot me a DM!

stone condor
#

Anyone have good project related to ai or machine learning with project report file?

#

@everyone

mint yoke
#

any quant here that successfully making money in trading with ML models

solemn kettle
upper shuttle
#

hello

thick vector
#

What's up everyone

whole stone
hexed cloak
#

Good morning!

naive crag
#

Hello 😊

opal ferry
#

Hello

burnt stream
#

.

unreal storm
#

Hello guys

trim igloo
#

Hello All..

shy mulch
#

hi

whole bane
#

hi

ionic lintel
#

hi

viral adder
#

hi

vital root
#

i'm a bit confused, is there a dedicated guide on how to submit in kaggle's competitions?
because it looks like writing a notebook is required, but my project is a mobile app and it is pushed to a github repo, can't i just submit the link to it ?

white blaze
#

Hello

chrome rover
harsh halo
#

Hello there

quartz basin
#

Hello,kaggle

night pasture
#

hi

lean stag
#

Hii

olive quartz
#

I’m looking for people who are interested in building something together — could be a side project, SaaS, AI tool, open-source idea, startup experiment, or literally any cool problem worth solving.

Skill level doesn’t matter as much as curiosity, consistency, and willingness to learn/build together.

If anyone’s up for collaborating, brainstorming, or shipping something interesting, DM me or reply here 🚀

#

If anyone already has an idea in mind, wants to work on something interesting, or is even planning for a hackathon, I’d love to join and collaborate

olive quartz
violet shadow
#

Back on track with Machine Learning & Data Science 🚀

Here’s my latest notebook about Diabetes Prediction using Machine Learning, where I went through the full workflow:

  • EDA
  • Handling null values
  • Handling duplicates
  • Handling outliers
  • Feature selection
  • Handling imbalanced data
  • Scaling features
  • Making ML models

You can read the notebook to understand every step I took during the project and the techniques I used throughout the analysis.

Check out the notebook here:
https://www.kaggle.com/code/minaremon39/diabetes-prediction

Also, feel free to follow me on Kaggle to keep up with future updates for this notebook and upcoming projects 📈

Waiting to get your feedback 🖤🖤

opal shard
#

How long did it take you to win your first kaggle competition by placing in top 10?

vital root
opal ferry
#

Hello beautiful people

uncut hazel
glacial temple
#

Hey, if i have a competition at my collage , and its topic is the bigmart sales one, any advice to lower error to 0.38 ? Our error now is 0.4

opal ferry
plain roost
glacial temple
#

I calc the error with MAE

#

Is it actually achievable, i mean the 0.38

opal ferry
#

Lol

#

Its ok

glacial temple
plain roost
opal ferry
glacial temple
#

I used catboostregressor

opal ferry
glacial temple
opal ferry
#

Models*

glacial temple
#

I did but actually cat boost is the best here

#

There is a method called stacking but actually it doesnt help much

#

Like merging two different models within a one model that learns to pick which one to use per point ig

opal ferry
#

Umm

#

Tune the model

glacial temple
opal ferry
#

Do you know how to tune models ?

glacial temple
opal ferry
glacial temple
wooden sleet
#

sorry to jump into the convo :3
but did u use optuna for tuning??? if not then try it

opal ferry
glacial temple
opal ferry
#

It will increade speed by 60-70%

#

Increase*

glacial temple
#

Ill try this , thankss all

rich haloBOT
#
progi_13777 has been warned

Reason: Posted an invite

lilac canyon
#

Hey

steel bison
#

i really liked the akkadian to english translation. but doing all the training myself with no team sucked. dont see any comps that particularly interest me atm.

#

cool community

rich haloBOT
#
enemy_1201 has been warned

Reason: Posted an invite

wet narwhal
wet narwhal
inland pagoda
#

Anyone here with crypto difficulties or any related problem

tropic stump
#

how do they manage to moderate the #1145832920049786880 ? do they have mods for every single language or wtf

rigid summit
#

hiiiiiiiiiiiii everynyan

#

im newwww

#

suppp

twin ore
#

Hello I am a Game Dev + Architecture Designer. I have done work with Drone making and i am looking forward to this course

plain roost
rigid summit
#

I'm new on kaggle too

#

just trying to learn

plain roost
rancid spear
#

hi

late tinsel
#

hi

elfin bolt
#

hiii , i'm a btech second yr student pursuing data science , trying to learn things

#

I’m new to Kaggle and Discord too😅

strong portal
#

same here

modest prawn
#

gmgm all

cerulean ravine
#

Hi , here to start 5-Day AI Agents Intensive Course with Google

pastel sedge
gray zealot
#

Hi 😊
I’m a B.Tech CSE AI & Robotics student from India. I’m interested in AI, Python, and app development. I enjoy connecting with students who are also learning technology and machine learning.

cyan ingot
gray zealot
#

Bro Currently I am Creating new projects only Not taking a Single Rupee from Anyone I have create 4 to 5 projects as A freelancer But Currently I am Investing My Time on My App/Project

#

So I am currently Using My money To devlop My app After App Devlopment And Launced in Market I will earn upto 10 to 50 Lakhs Per month with only user upto 1 Lakh Not more than that

#

My Aim is to achieve 1 Cr to 5 Cr Per Annum and this will be achieved by December 2027

uncut breach
#

Hey everyone!!

delicate kite
floral pulsar
#

Hi, I'm new here and just started, Happy to connect.

nova sapphire
#

Hii Everyone, I wanted to join a team we are just 2 members right now who are ready to build something great and give it all we got!

oblique mountain
#

Also, anyone play EVE online?

fast solar
#

Hi, This side Aayush

meager cradle
#

Hello

nova sapphire
plain roost
nova sapphire
plain roost
nova sapphire
#

SALIK-JAVID

#

Imran0511

elfin wasp
#

Hey everyone

nova sapphire
plain roost
nova sapphire
plain roost
unreal hinge
#

is it good if a middle schooler is competing

#

and getting above 75%

#

cuz i dont know the level of most people here

#

i belive it is good

#

to clarrify 75% percentile

topaz spoke
#

hello

odd terrace
#

hi

nova mirage
#

You guys are non secular

#

@everyone And everyone in this server are shameless

hallow roost
#

hy everyone , i am new here.

charred pawn
#

🚨 STOP sending boring CVs.

If you want your dream career, you need to stop using basic AI and start using the STAR Method.

I made a quick reel showing:
✅ How to turn a "Boring CV" into a "Selected" CV.
✅ The prompt that Senior Recruiters actually love.
✅ The #1 mistake that gets you rejected immediately.

Get the prompt here:
https://www.instagram.com/reel/DYOrbw3zcxI/?igsh=MTAzdTg0dzQ1dTFnbg==

somber verge
#

Hi new member here

plain roost
rose leaf
#

ok

plush rain
#

Okay

#

Hmm

#

Time to win

grave drift
#

hey guys

#

: )

old field
dawn geyser
#

👋 Hello everyone!

I’m Ansh Dham, currently an undergraduate at IIT Patna. My interests span across data science, machine learning, and building impactful projects that combine creativity with technical depth.

Beyond the technical side, I’m passionate about workflow automation, creative branding, and connecting with peers to collaborate on meaningful projects.

Excited to be here, to learn, share, and grow together with all of you. 🚀

dense pivot
#

I am eager to know what you guys think about AI? What is AI and Not AI? What do you think about the use cases and which AI Technology is revolutionizing the way we live?