Pytorch työt

The project centres on building a production-ready text-classification pipeline that leverages modern deep-learning techniques. I have a labelled dataset and need end-to-end code that ingests the text, handles cleaning and tokenisation, and trains an accurate classifier. Python is the preferred language; using PyTorch, TensorFlow or another mainstream framework is fine as long as the solution is reproducible and easy to extend. Key deliverables: • Well-commented source code (data loading, model, training loop, evaluation) • Clear instructions to run training on a fresh machine (README or notebook) • Metrics report showing accuracy, precision, recall and F1 on a held-out set • Exported model weights and a small inference script or API endpoint for batch pre...

...stack/Python developer. You will be helping us create and implement the following: • An adoption roadmap that ties specific AI capabilities to each stage of our workflow and project milestones, showing where automation, prediction, or generative content delivers the most value. • A reasoned “why this, not that” selection of tools—think Hugging Face transformers versus OpenAI GPT-4, TensorFlow or PyTorch for model training, spaCy for NLP, Vision APIs for image tasks—plus rapid prototypes that prove the choice. • Drop-in reference implementations or micro-services that slot straight into our existing Node/Express back end. • Plain-language docs and two-minute screen-share videos so community groups, national NGOs, and global partne...

I need a production-ready object detection model built, trained, and packaged so it runs smoothly on iOS and Android devices, in a modern web browser, and as a lightweight desktop application. The same model should power every platform to keep accuracy and behaviour consistent. You are free to choose the framework you are most comfortable with—TensorFlow, PyTorch, YOLOv8, Detectron2, or another proven library—as long as the final artefacts meet these requirements: • Mobile: optimised builds (e.g. TensorFlow Lite, Core ML, or ONNX) that hit realtime speeds on mid-range phones. • Web: WebAssembly/WebGL or implementation that loads in under three seconds on a standard connection. • Desktop: a small executable or Python app with GPU support when availab...

...levels—Admin, Standard user and Guest—each with appropriate permissions for running detections, reviewing results and managing data. Please structure the code so that REST endpoints are cleanly separated; this will let me expose the following Android-ready APIs later on: live-video analysis, image-file analysis and retrieval of disease-history logs. Deliverables • Python inference engine (TensorFlow/PyTorch + OpenCV acceptable) optimised for Raspberry Pi 5 • Django project with the described role system, templates and REST endpoints • Model-training notebook or script plus labeled dataset reference • Setup script or Dockerfile for one-step deployment on a fresh Pi • Brief README covering install, usage and endpoint documentation Ac...

...messages. We need a strong developer to implement the end-to-end pipeline: dataset ingestion, synthetic instruction generation, LoRA fine-tuning, validation harness, and local deployment with an OpenAI-compatible API. Target compute is on-prem GPU hardware (DGX Spark class). Optional experiments can run on Google Vertex AI credits. [Required Skills] - LLM fine-tuning: LoRA/QLoRA, quantization, PyTorch, Transformers, Unsloth preferred - Python data pipelines: JSONL, filtering, reproducibility - Mirth Connect (NextGen Connect): channel XML and Rhino JavaScript transformers - Healthcare interoperability: HL7 v2 (PID, PV1, OBR, OBX), FHIR basics - Docker, Linux, GPU environment setup - Bonus: vLLM, OpenAI-compatible serving, Google Vertex AI [Start Plan] Phase 1 is a vertical slic...

...rescuers can be dispatched quickly. I’m flexible about the imagery source—NASA, ESA, Google Earth, or any other free feed is fine as long as it delivers cloud-free, high-resolution scenes on a daily cadence. You are welcome to mix sources when one is fresher than another. The detector has to work at desert scale, so please build it with an established computer-vision framework (e.g., TensorFlow, PyTorch, YOLO, or a similarly robust model) and output the findings in both human-readable (an image with bounding boxes or a simple web map) and machine-readable form (CSV/GeoJSON with lat/long, time stamp, confidence score). Once I can point the script to a new polygon and receive a list of car and truck coordinates every 24 hours—fully automated, no manual clicks&m...

Rekisteröidy tai kirjaudu sisään nähdäksesi tiedot.

...intensity and volume to suit beginner through intermediate levels, and outputs a structured weekly routine (exercises, sets, reps, rest, and optional equipment notes). The system should justify its choices in plain language so the end user understands the reasoning behind each exercise block. I’d like the core to be written in Python, preferably leveraging familiar ML libraries such as TensorFlow, PyTorch, or scikit-learn—happy to hear your recommendation. Data sources for exercise selection can come from public, reputable fitness datasets or a curated ruleset you supply. The end result should be easy for me to integrate into a simple web or mobile front end. Deliverables • Clean, well-commented source code for the workout-plan generation engine • A ...

...blinking, lip-sync errors, or artifacts). * Uniform/Badge Recognition: Detect if the person is wearing a police uniform or showing a badge (using object detection like YOLO). * Real-Time Risk Dashboard: * A simple UI that displays a "Trust Score." If the score drops below a threshold, it shows a "SCAM ALERT" warning. Preferred Tech Stack: * Language: Python * ML Frameworks: TensorFlow / PyTorch / Keras * Computer Vision: OpenCV, MediaPipe * NLP: Hugging Face Transformers (BERT/RoBERTa for intent classification) * Interface: Streamlit or Flask (for the demo dashboard) Deliverables: * Source Code (well-commented). * A file for easy installation. * A short demo video showing the system detecting a scam attempt from a sample video file. * Document...

...caching, and Gunicorn as the application server for production; - Inventory Prediction and Purchasing Optimization Built predictive models using Prophet, XGBoost, LSTM, RNN, and PyTorch to forecast inventory levels and identify the optimal reorder point; Predicted future expenditures based on historical consumption, reducing costs from unnecessary purchases and preventing stockouts; Visualized results in Power BI to facilitate decision-making. - Financial Prediction by Department Built a predictive model to estimate revenue and costs per department for the following month; Utilized XGBoost, Prophet, LSTM, RNN, and PyTorch to model different financial behavior patterns; Implemented a data pipeline with PostgreSQL and ELT processes, ensuring efficiency in processing and in...

...raw mobile photos and automatically clean noise, sharpen details, fix color and convert them to 3d printable files. You’ll choose or design the model, train or fine-tune it, then wrap everything in a lightweight API that my mobile team can call in real time (on-device when feasible, cloud fallback when not). You should be completely comfortable with OpenCV plus deep-learning frameworks such as PyTorch or TensorFlow, and you know the trade-offs between traditional filters, GAN-based approaches, and modern super-resolution networks. Experience packaging models for CoreML, TensorFlow Lite or similar mobile runtimes will set you apart. I’m most interested in seeing what you’ve already shipped, so please include past work that proves you can take an image-processin...

...abstract this behind a clean API layer so either provider can be swapped in later. • Send confirmation and reminder messages automatically, with hooks for SMS and email. Tech preferences I am most comfortable reviewing code written in Python, and I already run Docker in production, so please containerise the service. You are welcome to use libraries such as spaCy, Rasa, Dialogflow, TensorFlow or PyTorch—choose what fits the problem, document your reasoning, and make installation reproducible with a single script. What I will accept as “done” 1. A Docker-ised project that I can spin up locally, connect to my calendar, and watch it schedule a demo meeting end-to-end. 2. Clean, commented source code plus a concise README detailing setup, environment ...

...collaboratively train a powerful intrusion-detection model without ever sharing raw traffic logs. Each embassy keeps its packets local, trains an IDS in Python, then ships only the updated weights to a permissioned blockchain that averages them and sends back the enhanced parameters. Key points I need turned into code: • Dual-framework training loops: the local agent must be able to switch between PyTorch and TensorFlow so each site runs whichever stack its engineers prefer. • Attack coverage baked into the dataset pipeline: the IDS must correctly classify Malware, Phishing, and DDoS patterns. Synthetic generation for edge cases is fine as long as the final validation shows clear detection rates for all three. • Smart-contract aggregator: the chain’s sole...

I'm looking for an experienced Python teacher specializing in advanced programming and machine learning. Essential requirements: - Expertise in Python, especially in machine learning - Ability to explain complex concepts clearly - Experience with popular ML frameworks (e.g., TensorFlow, PyTorch) - Flexible scheduling to accommodate my learning pace If you have a strong background in these areas, please apply!

I am rolling out a trio of inter-linked prototypes that tackle pressing urban and environmental challenges while showcasing cutting-edge AI, blockchain, and IoT engineering. First, the AI-driven Smart Traffic System must ingest multi-sensor feeds, apply real-time reinforcement lea...comparing baseline vs. prototype performance Acceptance criteria – Source compiles and runs without modification on the stated platform – ML models reproduce stated metrics using provided data – Voting demo passes independent audit script – Traffic simulation shows ≥ 15 % average wait-time reduction If you need data samples, CAD files, or clarifications on preferred libraries (TensorFlow, PyTorch, ROS2, Hardhat, ), let me know early so I can share the assets and keep...

...Record results, visualise them clearly, and explain the “why” behind the numbers • Iterate on your findings until we have a result worth integrating into production-grade code What you should bring Your application only needs to highlight relevant experience, but that experience must show strong analytical and problem-solving skills. If you are already comfortable in Python, R, TensorFlow, PyTorch or similar stacks, say so—those tools are the basis of most of our work, though deep mastery is less important than the ability to dissect a problem and reason through solutions. Deliverables 1. Reproducible notebooks or scripts for each experiment 2. A concise report explaining methodology, results, and next steps 3. A short hand-off call or recorded...

I’m running a research-grade study on how different transfer-learning strategies affect deep-learning image-classification accuracy and efficiency, and I’d like a PyTorch expert to help me turn the experimental plan into clean, reproducible code and an insightful report. Here’s what I need from you: • A well-structured PyTorch pipeline that can load a few supplied image datasets, swap in multiple pretrained backbones (e.g., ResNet, EfficientNet, ViT) and fine-tune them under several common strategies (feature extractor, full fine-tune, layer-freezing variants). • Training scripts or notebooks with clear logging of hyper-parameters, metrics, and run times so results are comparable across experiments. • A concise written analysis summa...

...and outputs clean, artifact-free H.264/H.265 versions at multiple resolutions. It should be efficient enough for batch processing on a single workstation but scale to a GPU server when required. • Recommend (and set up) an AI model suited to my animation goals, then wire it into the pipeline so I can move seamlessly from encoded footage to AI-augmented renders. Whether you lean on TensorFlow, PyTorch, or OpenCV is up to you—just justify the choice and document how to reproduce it on my side. • Provide concise documentation: command-line flags, filter graphs, model checkpoints, and any tuning parameters. I need to understand why each setting was chosen and how to tweak it for future projects. Acceptance criteria 1. A working FFmpeg script or shell command ...

I need a developer who can take this idea from concept to a working, browser-based product. The core objective is to enrol users with a unique “voice fing...prior to recording. Acceptance criteria • Enrolment and verification must run end-to-end in under five seconds on a typical broadband connection. • Equal Error Rate (EER) ≤ 5 % on a public speaker dataset or a comparable internal test set. • Clear documentation (setup, model training pipeline, API endpoints) plus a Dockerised deployment script. Tech stack is open, but Python (TensorFlow/PyTorch), Node.js, or Rust are all fine as long as you can justify the choice and meet the performance targets. Please outline your proposed approach, libraries you favour (Kaldi, SpeechBrain, etc.), and any simi...

...objective is simple: turn time-consuming manual work into an intelligent process that runs on its own. Because I don’t yet have data or legacy tools in place, we’ll define the workflow together, identify what information needs to be gathered, and then create the logic, models, or rule-based components that best fit. You’ll have full freedom to choose the right libraries—whether that’s TensorFlow, PyTorch, scikit-learn, spaCy, or a leaner approach using pure Python—so long as the end result is reliable, maintainable, and easy for me to extend later. Deliverables 1. Brief discovery call and written workflow outline 2. Prototype demonstrating successful end-to-end automation on sample data you create 3. Clean, well-commented Python code pl...

...I provide the raw market movements, the system should learn patterns and continuously refine its accuracy. Core deliverables: • Predict stock trends ahead of time • Analyze past market movements in an interactive timeline or chart view • Push real-time alerts the moment the model detects a significant event or deviation You are free to choose the most effective stack—TensorFlow or PyTorch for the model, and React Native, Flutter, or another proven framework for the apps—as long as the final build passes App Store and Google Play review. The AI must run efficiently server-side, expose a secure API to the app, and allow me to retrain or fine-tune the model with new data whenever needed. Please outline your proposed architecture, training work...

... The model must learn from three separate data streams—historical wheel outcomes, player betting behaviour logs, and detailed wheel-mechanics data—so it can pick up hidden biases and temporal patterns rather than relying on simple probability tables. Your task is to deliver a trained model plus everything I need to keep it learning as new data arrives. I’m comfortable with Python, TensorFlow/PyTorch, or a lightweight C++/C# inference library; whichever you choose, just expose a clean API that my front-end can call before each spin and return ranked predictions in well under 100 ms. Key deliverables: • End-to-end training pipeline (data cleaning, feature engineering, model training, validation) • Inference module or REST/gRPC service ready to drop...

...Evolutionary Test-Time Compute approach outlined in Jeremy Berman’s posts ( and follow-up). My main goal is stronger generalization, not just memorized accuracy, so the solution must show consistent gains across the public and hidden splits. Everything will run in Python. You can pick whatever supporting libraries you like (PyTorch, JAX, NumPy, Hugging Face, etc.) as long as setup stays lightweight and the code is clearly documented. Core scope • Implement evolutionary prompt search: generate, mutate, rank, and select prompts on-the-fly against ARC tasks. • Automate evaluation: scoring script should mirror ARC’s official rubric so results are directly comparable to the leaderboard. • Track sample efficiency: log compute

...report or dashboard, the assignment requires a fully realised project: working code, clear data-processing steps and concise documentation that I can present and defend. Here is what I need from you: • End-to-end pipeline: data sourcing or simulated dataset, cleaning and transformation, model training, evaluation and basic BI-style insights. • Well-commented Python or R notebooks (TensorFlow / PyTorch welcome) plus any auxiliary scripts. • A short explanatory document (≈5 pages) outlining objectives, methodology, results and how the solution fits within a Business Intelligence context. • Visual artefact: simple dashboard or set of plotted metrics that showcases key findings to a non-technical audience. Deep learning per visione artificiale; &bul...

Lead AI / Fullstack Engineer — Project "AZIZA" (Voice-to-Voice AI) ​Project Name: AZIZA Format: Project-based / Remote (with access to local GPU clusters) Tech Stack: PersonaPlex (Moshi-based architecture), PyTorch, TensorRT-LLM, FastAPI, WebRTC, Telegram Mini App (TMA). Hardware Location: Uzbekistan & Kazakhstan (TAS-IX), clusters powered by NVIDIA RTX 4090. ​Project Overview ​AZIZA is an innovative multimodal "Speech-to-Speech" (S2S) ecosystem designed to simulate natural human interaction. We are building an AI assistant that seamlessly transitions between roles: an expert tutor (Chemistry, History, Biology), an empathetic companion, and a simultaneous translator. By processing audio tokens directly, the system achieves unprecedented interaction spee...

I have collected numerical manufacturing data and need a working large-language-model approach that can accurately predict our experimental outcomes so we can cut material waste on the shop floor. Here’s what I need from you: • Build and train the LLM-powered model entirely in Python (TensorFlow, PyTorch, or similar—whatever you prefer). • Use my numerical datasets as the sole training source; they cover typical manufacturing readings and KPIs. • Validate the model’s performance with clear metrics and a concise report that shows how well the predictions match real results. • Supply clean, well-commented code plus a brief read-me so my in-house engineers can rerun or fine-tune the model later. Acceptance is straightforward: when you...

...comfortable with deep-learning pipelines and modern web development. First, you will fine-tune a large language model so it can deliver domain-specific answers through an API. Then you will create an e-commerce site that sells our digital products while showcasing the model’s capabilities. The model work involves preparing the dataset, selecting an appropriate base model, running the training (PyTorch or Python are fine), and exposing the result through a secure REST or GraphQL endpoint. Clear documentation on hyper-parameters, versioning, and how to reproduce the run is essential. For the website, I want a clean, responsive storefront with a shopping cart, payment gateway integration, and an admin dashboard for product and content management. Whether you choose React + N...

I need a compact electron vite Python-based pipeline that can watch hours of raw footage and automatically carve it into clean rushes. The idea is simple: detect every meaningful scene change or pronounced camera movement, mark the in- and out-points, and hand back an ordered timeline I can drop straight into my editing suite. My tool stack is already defined: PyTorch Video for the deep-learning backbone, PyDetect for proven vision utilities, and OpenCV for the faster classical operations. If you prefer to wrap parts of the workflow in ffmpeg or similar, that’s fine as long as the core logic stays in Python and can be called through a clear API endpoint. Scope • Analyse full-resolution clips, spot scene boundaries and camera pans/tilts/zooms with high recall. •...

...Natural Language Processing. My goal is to understand core NLP concepts, build a small portfolio of projects, and feel confident discussing them in interviews. I already code a little in Python but have no formal NLP background. I’d like to cover the foundations—tokenisation, embeddings, transformers—then progress to hands-on mini-projects using libraries such as spaCy, Hugging Face, TensorFlow or PyTorch. Guidance on best practices for data preprocessing, model selection, evaluation and error analysis is essential, along with tips on how to present this work in a résumé or GitHub repo. Deliverables I have in mind: • A personalised learning roadmap with milestones • Live or recorded sessions focused on theory followed by coding walk-t...

...interface quality, and overall fit. Once I’ve tested your demo, I’ll outline the domain specifics and grant data access. The engagement will revolve around refining your existing model, exposing a clean REST or GraphQL API, and deploying the solution to my cloud environment (AWS or GCP). Solid accuracy, low latency, and well-documented Python code that leverages libraries such as Hugging Face, PyTorch, TensorFlow, or LangChain are all critical. Deliverables: • A publicly accessible link to your working NLP demo • Brief but clear architecture and tech-stack summary • A step-by-step plan to adapt the model to my use case • Final integrated solution with deployment and usage documentation I’ll review submissions on a rolling basis and...

I need an end-to-end neural rendering pipeline that can faithfully reconstruct low-light scenes with NeRF. All training will be image-based and must run entirely in PyTorch. You are free to choose or combine publicly available night-time or low-illumination datasets, as long as the licence allows research use and you clearly cite the source. Here is what I expect at delivery: • A reproducible PyTorch implementation (data loader, network, training loop, rendering code). • Training and inference scripts that run on a single GPU and are easy to adapt to new datasets. • Built-in evaluation that logs PSNR, SSIM and LPIPS after every validation epoch and prints a final summary table. • A short report (markdown or PDF) discussing dataset choice, hyper-pa...

I want to build a full, low-light-oriented NeRF pipeline whose sole goal is to lift image quality. The core of the work is on the model architecture itself; I am happy to keep pre- and post-processing simple as long as the neural radiance field you design excels in dim scenes. What I need from you: • A PyTorch (or JAX/TensorFlow if you strongly prefer) implementation of a NeRF variant tuned for low-illumination data. • Training and inference scripts that I can run on a single-GPU workstation. • A concise evaluation notebook that reports PSNR, SSIM and LPIPS on the validation set so I can judge improvements immediately. • Clear instructions on how to add new scenes or fine-tune further. The work is finished when I can clone the repo, point it at my datase...

...later; please design the pipeline with those hooks in mind. Preferred flow 1. Upload or API push of multiple images. 2. Server chooses or lets me choose an alternative viewpoint/crop (“camera cut”). 3. AI engine processes, saves, and returns one or more high-res JPEGs per source image. 4. Dashboard shows progress, queues, and download links. Tech notes Feel free to lean on Python, TensorFlow/PyTorch, OpenCV, or similar libraries, plus a modern front end (React or Vue) and cloud storage (S3, GCS, or equivalent). Clean, well-commented source code, deployment scripts, and concise setup documentation must accompany the final hand-off. Deliverables (acceptance criteria) • Web portal deployed on a test server and accessible via URL • Batch upload handling at ...

...hands-on computer-vision engineer to turn raw sports footage into structured insight. The system you build will ingest live or recorded matches, lock onto the ball and key players, keep them centered with an intelligent zoom, and raise flags the moment a goal is scored or a significant positional change occurs. I expect you to stitch together proven object-detection or tracking APIs with your own PyTorch, TensorFlow, or OpenCV code so we hit production-grade accuracy without reinventing wheels. The pipeline should: • track players, the ball, and other relevant objects frame-by-frame, • adjust the crop dynamically so the action stays in focus, • recognise high-value events such as goals or point scoring, as well as nuanced player movements and formations, ...

...Troubleshoot and resolve technical issues - Stay updated with the latest industry trends and technologies Requirements: - Proven experience as a Full Stack Developer - Strong knowledge of front-end technologies (React, Angular, Vue, etc.) - Proficiency in back-end frameworks (Node.js, Django, Flask, etc.) - Experience with databases (SQL, NoSQL) - Solid understanding of AI/ML frameworks (TensorFlow, PyTorch, etc.) - Good problem-solving skills and attention to detail - Ability to work independently and as part of a team - Excellent communication skills Preferred: - Prior experience with long-term projects - Familiarity with cloud services (AWS, Azure, GCP) - Knowledge of DevOps practices What we offer: - Long-term collaboration - Flexible working hours - Competitive compensati...

...CRM, or recommendation systems). Conduct feasibility assessments, risk evaluations, and proof-of-concept development to accelerate time-to-market. Provide technical leadership and mentorship to ensure high-quality, maintainable code and architecture. Required Qualifications 8+ years of experience as an AI Architect or similar role, with proven expertise in AI/ML technologies (e.g., TensorFlow, PyTorch, GPT models, RAG systems). Strong background in bridging business needs to technical implementations, demonstrated through successful projects with short development cycles. Proficiency in cloud platforms (AWS, Azure, GCP) for AI deployment, MLOps, and scalable architectures. Experience in data engineering, NLP, computer vision, or predictive modeling relevant to business applicat...

...files—and I have already completed data cleaning, feature extraction, and normalization, so you can dive straight into modelling. Your task is to craft, tune, and ensemble models that out-perform the current leaderboard benchmark, then package everything into a fully reproducible training notebook and a submission-ready inference script. Python with scikit-learn, XGBoost, LightGBM, CatBoost, PyTorch, or TensorFlow are all welcome; pick the stack you believe will squeeze out the highest score. Deliverables • Commented training notebook(s) demonstrating data pipeline, model selection, tuning, and validation • Stand-alone inference notebook/script ready for Kaggle submission • README summarising architecture choices, hyper-parameters, CV strategy, an...

... • Flag high-risk clients to underwriters with an explanation score so they can override or request extra documentation. • Present all of this via a REST/JSON API that my current React front end can call, and expose a lightweight Python-based admin dashboard where underwriting managers can adjust thresholds and retrain models on fresh data. I work mainly with a Python solution (TensorFlow, PyTorch, or scikit-learn—whichever you feel is best) will plug in cleanly. The customer-facing chatbot or form assistant can be built with a service such as Rasa or a fine-tuned OpenAI model as long as it routes seamlessly into the same risk engine. Acceptance criteria 1. A reproducible training pipeline with clear data-schema docs. 2. Dockerised API that runs locally an...

...component that plugs directly into our existing order-processing software and reliably reads supplier purchase orders that arrive as JPEG scans. The engine only has to recognise English, but accuracy must be high enough to extract line-item tables, dates, totals and supplier IDs without manual correction. You are free to build on Tesseract, OpenCV, or a deep-learning stack such as TensorFlow or PyTorch, as long as the final module returns structured JSON we can map to our database. Deliverables I expect: • Source code with clear build instructions • A small training/validation dataset plus instructions for expanding it • API documentation and a quick demo script • Brief report on achieved accuracy and how to retrain the model if vendor layouts change ...

I have a closed-source loader that ships a set of PyTorch-based models stored in obfuscated “.bin” containers. They are not the usual TorchScript or ONNX packages; the application unwraps and feeds them directly to LibTorch at runtime. Your task is to reverse whatever protection the loader applies so I end up with usable, stand-alone TorchScript (.pt) files or another format that can be loaded with standard PyTorch APIs. What I know so far What I need from you 1. Analyse the loader binary, identify the decryption / unpacking routine, and reconstruct the original weights and network graph. 2. Hand over clean model files that I can load in a vanilla PyTorch script and run inference without the proprietary loader. 3. Document the steps or tooling you used we...

...building a web-based application that lets fashion brands and retailers drop a JPEG or PNG of any garment into the browser and instantly see that item rendered on a library of pre-trained human models in multiple poses and styles. The core of the product relies on machine-learning driven image-to-image translation, so I’m looking for someone who is equally comfortable training models (TensorFlow, PyTorch, or similar) and shaping a clean front-end experience (React, Vue, or plain JavaScript/HTML5). Here’s what I need you to deliver: • A fully functional web interface where users can upload JPEG or PNG files, preview results, and download the final composite. • A backend pipeline that handles preprocessing (masking, alignment, colour correction), inferenc...

I have a closed-source loader that ships a set of PyTorch-based models stored in obfuscated “.bin” containers. They are not the usual TorchScript or ONNX packages; the application unwraps and feeds them directly to LibTorch at runtime. Your task is to reverse whatever protection the loader applies so I end up with usable, stand-alone TorchScript (.pt) files or another format that can be loaded with standard PyTorch APIs. What I need from you 1. Analyse the loader binary, identify the decryption / unpacking routine, and reconstruct the original weights and network graph. 2. Hand over clean model files that I can load in a vanilla PyTorch script and run inference without the proprietary loader. 3. Document the steps or tooling you used well enough that I can...

...session slide decks, annotated notebooks or scripts, and any datasets you plan to use, shared ahead of each class. • Practical projects after every major module so learners leave with portfolio-worthy work. • Brief quizzes or checkpoints to confirm concepts have landed. • Recordings of the sessions for later review. Plan to weave in popular toolchains—think Python, scikit-learn, TensorFlow or PyTorch, Jupyter, plus the usual NLP and vision libraries—while keeping explanations clear, jargon-light, and immediately actionable. By the end, attendees should feel confident discussing model choices, training and evaluating them, and deploying simple prototypes. If you have run similar live cohorts before, let me know the approximate class size you&rsqu...

Project Title: Python ML Developer Needed for Generative Handwriting Tool (AI Style Transfer) Project Description: I am looking for a skilled Machine Learning developer to build a "Few-Shot Handwriting Synthesis" tool. The Co...Model: You have the creative freedom to choose the architecture (GANs, Diffusion Models, or Style Transfer). I am looking for the most realistic output. Build a Simple Interface: A basic local web page (Flask/Streamlit) to upload the image and generate text. Output Format: The system should generate high-resolution images (and preferably vector strokes if possible). Requirements: Experience with PyTorch/TensorFlow. Knowledge of Generative AI (GANs, Diffusion, or Transformer-based style transfer). Ability to work with Image Processing (OpenCV) to cle...

...& reporting – actionable insights, trend visualizations, and auto-generated reports that decision-makers can consume without data-science fluency. • System optimization & automation – feedback loops that tune system performance, resource allocation, and patches/updates based on the AI’s findings. Tech expectations Rapid prototyping and model experimentation can live in Python (TensorFlow, PyTorch, scikit-learn—whichever suits), while performance-critical inference or low-latency modules may be rewritten in C++. I’ll host on our existing infrastructure, so clean APIs, a modular codebase, and thorough documentation are essential. Deliverables 1. Trained models with reproducible training pipelines 2. Source code (Python & C++)...

...neural fusion layer; I’m happy to discuss which choice achieves the best balance of speed and robustness. The finished model must be benchmarked on four fronts that matter equally to me: Accuracy, Precision, Recall, and False-Alarm Rate. Because the project will eventually guard a sensitive site, I’m aiming for a solution that can move from laptop to field hardware without major rework. Python, PyTorch, and standard CV/RF libraries are expected; please keep your code clean, well-commented, and reproducible. Deliverables: • Trained YOLOv8 weights and training notebook • Trained RF-CNN model and training notebook • Fusion module with documented logic or architecture • End-to-end inference script that ingests live or recorded video plus I/Q da...

...justify the choice and keep dependencies light. The key deliverables are: 1. Fully functional one-time payment flow using Stripe. 2. AI-driven categorisation of each successful payment, stored in my data store. 3. Clear, step-by-step setup instructions so I can reproduce the configuration in staging and production. If you have previous examples of pairing Stripe with ML tools like TensorFlow, PyTorch, or even a SaaS NLP classifier, that would be great to see, but I mainly care that the final handoff is clean, tested, and documented....

...statistical analysis explaining performance differences, significance tests, and potential limitations suitable for the thesis methodology chapter. • Step-by-step write-up support: comments in code, inline LaTeX-ready tables/figures, and brief Zoom walkthroughs so I can defend each decision confidently. Tools currently in use: TensorFlow/Keras, scikit-learn, pandas, NumPy, Jupyter Lab. If you prefer PyTorch for certain parts, I’m flexible as long as the reasoning is sound and the results are reproducible. Timeline is tight—ideally the first working CNN model within two weeks—so proven experience delivering similar academic projects is a must. Please share one example of a previous thesis or peer-reviewed paper you helped shape, especially if it involved...

...understand a prompt such as “Show me the expected CPU, memory and disk utilisation for the next 24 hours” and then: • craft the appropriate SQL query that pulls historic utilisation metrics (CPU, MEM, DISK) from my monitoring tables, • run or hand that query off to the existing pipeline, • feed the results into a prediction model (your choice of proven libraries like scikit-learn, TensorFlow, PyTorch, Prophet, etc.), • return a concise, human-readable forecast plus the raw numbers so we can visualise them in our dashboard. You will need to wire the ML inference step into the current codebase, expose it through the same API layer the LLM already uses, and document the new endpoint and model-training process. I am DB-agnostic for now—if ...

...into a recommended number of units b) Provide uncertainty handling (confidence bands or reliability rating) c) Must be explainable for non-technical stakeholders 5) Handover package a) Clean Python codebase + README b) Validation report (results, risks, limitations) c) “Plain English” summary for business stakeholders Tech Stack (Preferred) 1. Python 3.10+ 2. PyTorch 3. PINN tooling: DeepXDE or custom PINN in PyTorch 4. pandas, numpy, scikit-learn 5. PDF extraction: pdfplumber + (Camelot/Tabula if needed) Optional: MLflow or W&B for experiment tracking Optional: FastAPI or Streamlit for demo Required Experience (Non-Negotiable) Proven experience with time-series forecasting on messy real-world datasets Hands-on experience with Physics-Info...