Loraemplois

...j’attends : • Un rendu 100 % cartoon, cohérent sur l’ensemble de la scène. • Présence obligatoire d’un paysage et d’éléments architecturaux ; aucun autre type d’élément n’est requis. • Un fichier de workflow (.json) contenant tous les nœuds ComfyUI nécessaires : génération, upscale, contrôle de la palette, puis découpe en tuiles (par exemple 1024 px ou paramétrable). • Indication des modèles, LoRA ou checkpoints utilisés et des réglages clés (seed, sampler, dimensions, steps). • Brève documentation indiquant comment lancer le workflow, modifier le prompt, ajuster la taille finale e...

...dans le but d’établir ensuite une collaboration récurrente (plusieurs Reels par semaine). Éléments fournis : Scripts détaillés scène par scène Voix-off déjà enregistrées (fichiers WAV propres) Photos du speaker + fichier .safetensors LoRA du modèle du speaker (pour une génération ultra-réaliste et cohérente) Vidéo de référence complète (partie 1) pour le style, le rythme et la direction artistique Accès à notre compte Envato Elements Tâches principales : Génération du speaker en vidéo IA avec synchronisation labiale parfaite (lip sync) en exploitant le LoRA fourni Création et s&e...

...dans le but d’établir ensuite une collaboration récurrente (plusieurs Reels par semaine). Éléments fournis : Scripts détaillés scène par scène Voix-off déjà enregistrées (fichiers WAV propres) Photos du speaker + fichier .safetensors LoRA du modèle du speaker (pour une génération ultra-réaliste et cohérente) Vidéo de référence complète (partie 1) pour le style, le rythme et la direction artistique Accès à notre compte Envato Elements Tâches principales : Génération du speaker en vidéo IA avec synchronisation labiale parfaite (lip sync) en exploitant le LoRA fourni Création et s&e...

Le Label Abeille, marque de la société BEE ANGELS, commercialise un objet connecté qui permet de suivre à distance et en temps réel la santé des abeilles à travers différents capteurs (poids, température, humidité, géolocalisation, orientation...). Plus d'infos sur...qui permet de suivre à distance et en temps réel la santé des abeilles à travers différents capteurs (poids, température, humidité, géolocalisation, orientation...). Plus d'infos sur Nous sommes à la recherche d'un développeur freelance senior ou qualifié, afin d'optimiser un code existant sous microcontrôleur Microchip. Maîtriser impé...

Développement Python - Modbus Bonjour, Pour un nouveau projet, je dois récupérer le niveau d’un plan d’eau et l’envoyer à un bateau. Sur le bateau, un logiciel récupéra la valeur pour permettre un forage dirigé. La transmission radio devrait se faire avec des modems Atim ARM/868. Il nous manque toute la gestion des requêtes Modbus. Es...utilisant une librairie : • Lire la valeur du capteur, certainement en hexadécimal, • Traduire la trame Modbus et la mettre à disposition d’un PC sous format Excel, csv, ou autre format à définir, • Nous souhaitons faire une lecture à une fréquence entre 30s à 1 minute. Le programme devra pouvoir évolu...

I am building an AI companion platform and have already created character faces using MidJourney. I now want to transition to a more professional SDXL + LoRA pipeline to: • Improve realism • Maintain consistent character identity • Lock faces for long-term use • Build scalable influencer-ready assets I am NOT looking for someone to generate images for me. I am looking for expert consultation and workflow optimisation guidance.

...a high-speed USB hub, balanced power distribution, and impeccable signal integrity without exceeding the layer budget. Components to Integrate: Component Details Interface Luckfox Lyra Zero W ARM Cortex-A7 Linux SBC (65×30mm, 40-pin header) Direct solder or socket Teensy 4.1 ARM Cortex-M7 @600MHz audio DSP (61×18mm) 2.54mm headers SL2.1A USB Hub IC 4-port USB 2.0 hub Surface mount XIAO ESP32S3 LoRa module (21×18mm) 1.27mm headers STM32WB55 Module Dual-core ARM Cortex-M4/M0+ (Flipper brain) 8-pin header PyBoard v1.1 ARM Cortex-M4 backup controller 6-pin header GPS Module NEO-6M/8M 4-pin header CC1101 Module Sub-GHz radio 8-pin header ST25R3916 Module NFC/RFID 8-pin header OLED Displays (2x) 1.3" 128×64 I²C 4-pin headers Trackball Module BlackBerr...

...weeks Script parsing into structured shot lists using our training data. Image generation via FLUX Kontext Pro with real estate agent composited into scenes from reference photos. Video clip generation from approved frames via API. Voiceover via Minimax TTS. Approval UI for reviewing and regenerating each shot. Phase 2 — $5,000-$7,000, 6-8 weeks Migrate generation to RunPod with ComfyUI workflows. LoRA avatar training pipeline. Self-hosted lip sync. Batch processing. API bridge between storyboard app and ComfyUI. Phase 3 — $5,000-$8,000, 6-10 weeks Automated video assembly with transitions, captions, music. Premiere Pro XML export. Template system for different video types. Full pipeline orchestration with error recovery. Required: Production ComfyUI experience (show...

The project related stage 2 for integrating the Esp32 and LoRa module.

It is 2 stage project 'I have the source code available, i have the API dataset available. But definately need a NEW DNN training Dataset. simple UI . Total P...SMS gateway integration and push notification backend Secure API development and logging system Deliverable: Deployed ML model with working API endpoint and backend system. STAGE 2 — Embedded & Mobile Systems Engineer (LoRa + App + Hardware) Looking for an engineer experienced in: ESP8266/ESP32 (NodeMCU) firmware development SX1278 LoRa 433 MHz communication and mesh setup Sensor integration (analog/digital input to MCU) Android app development with GPS + map integration API integration between app and server Offline communication architecture Deliverable: Functional LoRa communication modul...

...technician to develop high-fidelity LoRA models for FLUX.1 [dev] and Wan 2.1. The goal is to achieve [insert goal: e.g., high character likeness / a specific artistic style] that remains consistent across both high-resolution images and video generation. ​Required Technical Expertise ​Model Experience: Proven experience training LoRAs for FLUX.1 (Dev/Schnell) and Wan 2.1 (specifically the 1.3B or 14B models). ​Tooling: Proficiency with Ostris AI-Toolkit, Musubi-tuner, or Kohya_ss. ​Dataset Management: Ability to curate and caption high-quality datasets (images for Flux, video clips for Wan). Experience with Danbooru/WD14 tagging or natural language captioning (LLM-based). ​Temporal Consistency: For Wan 2.1, you must be able to demonstrate that the LoRA maintains ident...

...E220-400T22S LoRa module. My next step is to merge these routines into one cohesive, ultra-low-power application on our custom PCB. Power optimisation is the top priority. I want the system to remain in STOP mode most of the time, waking through the LIS3DH interrupt and periodic RTC events. When active, each peripheral must communicate reliably without compromising the sleep budget. Key points • Integrate existing C-based drivers into a single project • Configure and test the STOP-mode strategy, ensuring wake-ups from both the LIS3DH and the RTC function every time. • Audit clocks, GPIO states, and peripheral gating to reach the lowest possible average current—battery life is critical. • Maintain interrupt safety and data buffering so RFID reads...

Veuillez vous inscrire ou vous connecter pour voir les détails.

...interfaces with multiple sensors for two separate hives. The device will be powered by a Solar panel and a Li-Ion/LiFePO4 battery. Core Components & Requirements: Microcontroller & Connectivity: Based on ESP32 with LoRaWAN capabilities. Preference: You can either design around a module (like ESP32-WROOM + RFM95) OR design a carrier board for a dev-board like the Heltec WiFi LoRa 32 V3 (to simplify assembly). Please specify your recommendation in your proposal. Sensors (Interfaces needed for 2 Hives): Weight: 2x Interfaces for Load Cells (4-wire). The PCB must integrate 2x HX711 ADC chips directly (or provide headers for breakout boards, though integrated chips are preferred for reliability). Audio: 2x Interfaces for external ...

...mehrerer Puppen über Wi-Fi und LoRa; OTA-Updates und zentrale Steuerung Mein aktuelles Setup nutzt SPI und I²C-Peripherie sowie erste Sensor-Breakouts. Nun brauche ich: 1. Optimierte Hardware: Schaltplan, Platinenlayout, saubere Strom­versorgung, ggf anpassung der Komponenten . 2. Firmware: stabile Echtzeit­logik für Audio, Heat-Control, Sensor­auswertung, Wi-Fi/LoRa-Mesh, OTA-Update­mechanismus. 3. Prototyp-Validierung: Aufbau von 1–2 neuen Boards, Tests (Reichweite, Akku­laufzeit, Temperaturprofil), ggf. CE-Vorklärung. 4. Dokumentation: Code-Kommentar, Flash-Anleitung, Hardware-Dossier, damit spätere Stückzahl­fertigung problemlos klappt. Ich arbeite gern iterativ: kurze Sprints, getestete Teilergebnis...

...PWM OUT Woking and control wheels 2. Make the joystick Inputs working and get the ADC inputs from Joystick (this i have already done , half rest needs to be continued) Digital Input work as E-STOP input 4. Make Proximuty sensor and Ultrasonic sensor working With this i am going to prepare a Robotics vehicle for my FARM For sparying purpose . Future Tasks : 1. Make the input working using LORA and INTERNET (may be SIM800 Module) Process : 1. Requirement document will be provided 2. design document shall be prepared and reviewed after that only code shall be done 3. Must Follow MISRA 2012 Standards code shall be sent for review and review comemnts hsall be implemented 5. code shall be archived in github NOTE : the code shall be implemnted on STM32 Controller , so the person ...

...and hashtags), extracts trend “DNA” (patterns rather than clones), and uses that intelligence to generate scripts, visuals, captions, CTAs, and hashtags optimized per platform. Video generation is pluggable and provider-agnostic, supporting Bring-Your-Own-Key (BYOK) integrations with third-party AI services and a self-hosted fallback pipeline to guarantee uptime and protect margins. A first-class LoRA Character System enables consistent, recognizable AI characters across multi-scene videos. Users can design characters inside the platform or import third-party LoRAs (e.g., from Civitai), manage versions, enforce licensing rules, and maintain visual identity through controlled prompts, weights, and seeds. The platform includes a robust scheduler, auto-posting connecto...

...and hashtags), extracts trend “DNA” (patterns rather than clones), and uses that intelligence to generate scripts, visuals, captions, CTAs, and hashtags optimized per platform. Video generation is pluggable and provider-agnostic, supporting Bring-Your-Own-Key (BYOK) integrations with third-party AI services and a self-hosted fallback pipeline to guarantee uptime and protect margins. A first-class LoRA Character System enables consistent, recognizable AI characters across multi-scene videos. Users can design characters inside the platform or import third-party LoRAs (e.g., from Civitai), manage versions, enforce licensing rules, and maintain visual identity through controlled prompts, weights, and seeds. The platform includes a robust scheduler, auto-posting connecto...

...Union offering high-value technical PLANNING - GERMAN STANDARD and concept services to clients across industry, logistics, hotels, and commercial facilities. We are looking for experienced freelancers to support our projects in the following areas: Areas of Expertise: RFID Access Solutions – industrial facilities, hotels, commercial buildings, machine authorization Radio & Antenna Concepts – LoRa, WLAN, LTE/5G indoor coverage, antenna layout and field planning Automation Concepts – PLC design, process control workflows (planning only, no on-site execution) Robotics Concepts – educational and industrial robot integration, process optimization What We Offer: Remote collaboration across the EU Clear project scope and deliverables Professional, lo...

...building an ecosystem of ESP32-S3-based modules and I’d like to bring a seasoned embedded developer on board for the long haul. Every module in the roadmap—starting with the very first one we will tackle together—must combine three pillars: LoRa communication (with robust AES-based encryption), a lightweight web panel served from the device itself, and an on-device GUI rendered on a small display. For the initial milestone you will deliver fully-tested, production-ready firmware that proves this trio works seamlessly on the S3: flawless encrypted LoRa packets, a responsive web panel for configuration/monitoring, and a smooth GUI (LVGL or comparable) sharing the same data layer. I’m working with ESP-IDF, but I’m open to your preferred toolchai...

I’m diving into Andrej Karpathy’s nanochat repo and want an expert to sit with ...to understand the codebase and to map out training methodologies so I can adapt the model to a domain like law or healthcare. Start your bid with the word tinychat so I know you’ve read this. If you have already cloned, modified, or trained nanochat, please highlight that. I’m also keen to hear about any other LLMs you’ve fine-tuned, which verticals you tackled, and what results you achieved (LoRA, PEFT, Hugging Face Transformers, RAG pipelines, etc.). Timing is flexible; we’ll coordinate a slot that works for both of us. We will also discuss next steps after the first call. Deliverable • One live, screen-shared session (~60 min) where we discuss nanochat...

...based in the European Union offering high-value technical PLANNING and concept services to clients across industry, logistics, hotels, and commercial facilities. We are looking for experienced freelancers to support our projects in the following areas: Areas of Expertise: RFID Access Solutions – industrial facilities, hotels, commercial buildings, machine authorization Radio & Antenna Concepts – LoRa, WLAN, LTE/5G indoor coverage, antenna layout and field planning Automation Concepts – PLC design, process control workflows (planning only, no on-site execution) Robotics Concepts – educational and industrial robot integration, process optimization What We Offer: Remote collaboration across the EU Clear project scope and deliverables Professional...

We are PKYDrip, a B2B company building a modular ...distributors. This is NOT consumer marketing. We focus on clear system logic, integration, and real-world use cases. This budget covers a trial package of 3 pages (2 solution pages + 1 product page). Scope (Month 1) Write 3 pages: Solution Page – Smart Water-Saving Irrigation for Agriculture (800–1200 words) Solution Page – Smart Water-Saving Irrigation for Greenhouse (800–1200 words) Product Page – LoRa Valve Controller (600–800 words) Deliverables Google Doc / Word Clear H1–H3 headings One round of revision included English only for now (more languages later) This is NOT consumer marketing content. Please do NOT submit AI-generated generic blog samples. We are looking for cle...

I need a LoRa AI that creates enchanting and attractive personas for online and social applications. Requirements: - Compatible with web and mobile/social apps - Functionality to create and manage AI personas - Enchanting and attractive persona characteristics Ideal skills and experience: - Expertise in AI development - Experience with social and online app integration - Strong design skills for persona creation - Prior work with user engagement and interaction Please share relevant portfolios and past experiences.

...domain-specific question-answering system is live, with a fine-tuned LoRA/SFT language model sitting in a private Hugging Face repo, served through a FastAPI micro-service and wrapped in a multi-agent pipeline (retriever + analyst). Everything ships in Docker, rolls out under Kubernetes via Helm, and already includes a basic monitoring and evaluation loop. Right now the bottleneck is accuracy—specifically the relevance of each answer. I have a fresh batch of curated domain data ready for ingestion, and I’d like that folded into the workflow along with any architecture or hyper-parameter tweaks that further sharpen responses. Key focus areas for today (hard deadline: 7 pm): • Integrate the new dataset, run a quick SFT/LoRA refresh, and redeploy the updat...

## Job Title: Expert Audit of Whisper Large-v3 LoRA Script – Code Review & Bug Fixes ### Job Description: I have developed a custom Python script for fine-tuning **Whisper Large-v3** using **LoRA (PEFT)** on the Czech Common Voice dataset. While the script runs, I am facing specific technical debt and logic issues that require a senior-level ML Engineer. **STRICT REQUIREMENT:** I am NOT looking for a new script or a "black-box" solution. I want to work with **my existing codebase**. The goal is to identify, explain, and fix the bugs within my logic so I can understand the underlying issues and continue developing this specific script. ### Current Challenges to Solve: 1. **Persistent Attention Mask Warning:** *"The attention mask is not set and ca...

Hypothesis: A constraint-aware, evidence-linked RAG architecture specifically optimized for Small Language Models (SLMs) can achieve legal reasoning accuracy and hallucination reduction comparable to or ...sentence embeddings + FAISS index. 3. Reranker — cross-encoder reranker (InLegalBERT fine-tuned) to re-rank top-k. 4. Constraint module — set of lightweight rule engines/filters enforcing legal constraints before generation. 5. Evidence formatter — organizes retrieved snippets + provenance into generator context. 6. Generator (SLM) — small seq2seq (FLAN-T5 small/TinyLlama / Phi-4) fine-tuned with LoRA/QLoRA to output multi-field evidence-linked answers. 7. Hallucination detector / verifier — checks claims against retrieved evidence; optionally runs a...

...report the data in real time. The core task is straightforward: bring together barcode reader, cameras, Infrared camera, GPS receiver (RTK), and ToF sensors so the platform can see, focus, and reliably decode barcodes as it moves—no manual intervention, no tethered laptop. Besides this, also take IR image of the scanned area. Data must be stored locally and also communicated to a server over wi-fi/LoRa/4G network, along with with precise GPS location. Push button - start and stop. This is a non-AI/ML project. Open source software only. No dependencies or additional costs of software. Project scope: 1. Integration of 2 barcode readers, 4 small cameras, 1 higher resolution camera, 1 IR camera, 4 geared motors for driving the wheels and synchronizing them; control of the sp...

...Mengirim data secara real-time ke dashboard berbasis web atau aplikasi Android sederhana, lengkap dengan notifikasi bila parameter di luar batas. • Menyediakan diagram kabel, skema PCB (jika diperlukan), kode program terkomentar, serta panduan instalasi langkah demi langkah agar saya bisa mereplikasi dan melakukan perawatan mandiri. • Memberi rekomendasi housing tahan cuaca, konektivitas (Wi-Fi, LoRa, atau GSM) yang sesuai dengan medan pertanian di Indonesia. Kriteria keberhasilan: – Semua sensor terbaca stabil minimal 7×24 jam tanpa reset manual. – Dashboard menampilkan grafik historis dan mampu mengekspor data CSV. – Notifikasi tersampaikan di ponsel dalam waktu <1 menit setelah ambang batas terlampaui. Saya terbuka berdiskus...

PCB 1: Board type: 2-layer PCB MCU: ESP32 WROOM Wireless: LoRa Ra-02 Power: • 230V AC input • Protection: Fuse, MOV, NTC • AC → HLK-PM01 (5V) • 5V → 3.3V LDO Schematic pages are created, GPIO assignments are mostly finalized, PCB placement is partially done, Routing is incomplete and needs professional cleanup, Special attention needed for mains isolation and grounding PCB 2: Temp Node ESP32 C3, LoRa , DS18B20

...member core team specializing in full-cycle IoT and embedded product help startups, researchers, and companies turn ideas into fully working, production-ready IoT products, from PCB and firmware to cloud, AI, and mobile Core Expertise Covers the Entire IoT Chain-> Embedded Firmware DevelopmentSTM32, ESP32, Arduino, nRF, PIC, Raspberry PiBare-metal & RTOS (FreeRTOS)BLE, Wi-Fi, LoRa, MQTT, CAN, SPI, I²C, UARTSensor integration, power optimization, OTA updates-> PCB Design & Hardware EngineeringSchematic + PCB design (EasyEDA, Altium, KiCad)Single-layer to multilayer boardsBOM optimization & DFM/DFA supportPrototype-ready Gerbers & manufacturing files-> Cloud, Backend & AI IntegrationAWS, Firebase, custom backend APIsIoT dashboards & device ma...

I am looking for an experienced PCB design engineer to review, validate, and complete PCB routing for an embedded IoT project already designed in EasyEDA Pro (Desktop). The schematic is largely complete, and initial PCB placement has been done. I need a professional to verify correctness and finish the PCB properly. Tools & Requirements • Must use EasyEDA Pro (Desktop) • Experie...am looking for an experienced PCB design engineer to review, validate, and complete PCB routing for an embedded IoT project already designed in EasyEDA Pro (Desktop). The schematic is largely complete, and initial PCB placement has been done. I need a professional to verify correctness and finish the PCB properly. Tools & Requirements • Must use EasyEDA Pro (Desktop) • Experience...

...adultos. El objetivo es desarrollar modelos visuales realistas, coherentes y escalables para su uso en distintas plataformas. Funciones del puesto Creación de modelos IA hiperrealistas (rostro, cuerpo, expresiones, estilos) Manejo de herramientas y programas de generación de imágenes y video con IA Configuración, ajuste y optimización de prompts avanzados Entrenamiento y ajuste de modelos (LoRA, fine-tuning u otros sistemas similares) Edición y postproducción de foto y video (retoque, corrección de errores, upscale, coherencia visual) Mantenimiento de una identidad visual consistente del modelo Generación de contenido variado (poses, outfits, escenarios, estilos) Optimización del flujo de trabajo para ...

...videos will use AI-generated voiceovers and clean, academic motion graphics to explain complex legal topics and exam-oriented answers. ​2. Brand Guidelines (Strict Adherence) ​Color Palette: Ivory (#FFFFF0) background, Navy Blue (#000080) for headers/primary text, Maroon (#800000) for highlighting key terms and case laws. ​Typography: Professional Serif fonts (e.g., Playfair Display, Cinzel, or Lora) for a "legal" feel. ​Aspect Ratio: 16:9 (YouTube Desktop) and 9:16 (YouTube Shorts). ​3. Scope of Work (Per Video) ​Script Adaptation: Taking a provided legal note (300–800 words) and adapting it into a conversational, engaging script. ​AI Voiceover: Generating a natural, professional, neutral-accent Indian-English voice (e.g., ElevenLabs or WellSaid Labs). No "r...

...videos will use AI-generated voiceovers and clean, academic motion graphics to explain complex legal topics and exam-oriented answers. ​2. Brand Guidelines (Strict Adherence) ​Color Palette: Ivory (#FFFFF0) background, Navy Blue (#000080) for headers/primary text, Maroon (#800000) for highlighting key terms and case laws. ​Typography: Professional Serif fonts (e.g., Playfair Display, Cinzel, or Lora) for a "legal" feel. ​Aspect Ratio: 16:9 (YouTube Desktop) and 9:16 (YouTube Shorts). ​3. Scope of Work (Per Video) ​Script Adaptation: Taking a provided legal note (300–800 words) and adapting it into a conversational, engaging script. ​AI Voiceover: Generating a natural, professional, neutral-accent Indian-English voice (e.g., ElevenLabs or WellSaid Labs). No "r...

...across various poses, environments, and lighting conditions while maintaining a high degree of anatomical realism. Key Responsibilities: LoRA Training: Create custom LoRAs to lock in specific facial features and body types for consistent character generation. High-Fidelity Outputs: Implement "Nano Banana," Grok-based insights, or the latest open-source tools to push the boundaries of realism and skin texture. Consistency: Ensure the model remains identical across different "scenes," including full-body shots and close-ups without clothing. Technical Requirements: Expertise in Stable Diffusion (Automatic1111 or ComfyUI). Proven experience in LoRA training (Kohya_ss or similar). Deep understanding of character-consistency techniques (IP-Adapter, F...

...member core team specializing in full-cycle IoT and embedded product help startups, researchers, and companies turn ideas into fully working, production-ready IoT products, from PCB and firmware to cloud, AI, and mobile Core Expertise Covers the Entire IoT Chain-> Embedded Firmware DevelopmentSTM32, ESP32, Arduino, nRF, PIC, Raspberry PiBare-metal & RTOS (FreeRTOS)BLE, Wi-Fi, LoRa, MQTT, CAN, SPI, I²C, UARTSensor integration, power optimization, OTA updates-> PCB Design & Hardware EngineeringSchematic + PCB design (EasyEDA, Altium, KiCad)Single-layer to multilayer boardsBOM optimization & DFM/DFA supportPrototype-ready Gerbers & manufacturing files-> Cloud, Backend & AI IntegrationAWS, Firebase, custom backend APIsIoT dashboards & device ma...

...SEO-optimized structure PDF viewer integration Newsletter subscription (e.g., Mailchimp) Social media links and Google Analytics User roles management (Admin, Author, Editor, Reviewer) Technical Preferences: Hosting: 10Web, Hostinger, or Bluehost Recommended Plugins: WPForms, PDF Embedder, RankMath/Yoast SEO, Mailchimp, User Role Editor Branding: Green, Brown, White color scheme; Open Sans & Lora fonts; clean, modern style Deliverables: Fully functional WordPress website with all pages and features Mobile-friendly and SEO-optimized design Secure submission forms and user-friendly backend Integration with newsletter and analytics Future Scope: Subdomains for additional agri services Premium subscription content AdSense / affiliate monetization Multilingual su...

...adultos. El objetivo es desarrollar modelos visuales realistas, coherentes y escalables para su uso en distintas plataformas. Funciones del puesto Creación de modelos IA hiperrealistas (rostro, cuerpo, expresiones, estilos) Manejo de herramientas y programas de generación de imágenes y video con IA Configuración, ajuste y optimización de prompts avanzados Entrenamiento y ajuste de modelos (LoRA, fine-tuning u otros sistemas similares) Edición y postproducción de foto y video (retoque, corrección de errores, upscale, coherencia visual) Mantenimiento de una identidad visual consistente del modelo Generación de contenido variado (poses, outfits, escenarios, estilos) Optimización del flujo de trabajo para ...

...member core team specializing in full-cycle IoT and embedded product help startups, researchers, and companies turn ideas into fully working, production-ready IoT products, from PCB and firmware to cloud, AI, and mobile Core Expertise Covers the Entire IoT Chain-> Embedded Firmware DevelopmentSTM32, ESP32, Arduino, nRF, PIC, Raspberry PiBare-metal & RTOS (FreeRTOS)BLE, Wi-Fi, LoRa, MQTT, CAN, SPI, I²C, UARTSensor integration, power optimization, OTA updates-> PCB Design & Hardware EngineeringSchematic + PCB design (EasyEDA, Altium, KiCad)Single-layer to multilayer boardsBOM optimization & DFM/DFA supportPrototype-ready Gerbers & manufacturing files-> Cloud, Backend & AI IntegrationAWS, Firebase, custom backend APIsIoT dashboards & device ma...

I need a fully-trained Z-Image Turbo LoRA that can convincingly reproduce the look of Julia Zagajska (reference photos on her Instagram: ). I’m happy for you to follow the RunPod-based workflow in the PromptingPixels tutorial () or any comparable setup, as long as the result is compatible with Stable Diffusion 1.5 and Automatic1111.

We are looking for an AI/ML engineer who can fine-tune an open-source image generation model and build an image-to-image pipeline. The system must handle pose-based inputs, clothing, character, and background control. We plan to use this in production, so speed, accuracy, and stability are important. You will help us: Build ...generation Maintain strong prompt + visual accuracy Produce deterministic results Deliver a clean inference API (REST or gRPC) Performance target: Image generation in about 5–8 seconds on a 4090 GPU. We want someone who can work with us long-term as we continue developing features like LoRAs, presets, and video generation later on. If you have experience with diffusion models, ControlNet, SDXL, LoRA training, or production deployment, we want to ...

We are looking for an AI/ML engineer who can fine-tune an open-source image generation model and build an image-to-image pipeline. The system must handle pose-based inputs, clothing, character, and background control. We plan to use this in production, so speed, accuracy, and stability are important. You will help us: Build ...generation Maintain strong prompt + visual accuracy Produce deterministic results Deliver a clean inference API (REST or gRPC) Performance target: Image generation in about 5–8 seconds on a 4090 GPU. We want someone who can work with us long-term as we continue developing features like LoRAs, presets, and video generation later on. If you have experience with diffusion models, ControlNet, SDXL, LoRA training, or production deployment, we want to ...

We are looking for an AI/ML engineer who can fine-tune an open-source image generation model and build an image-to-image pipeline. The system must handle pose-based inputs, clothing, character, and background control. We plan to use this in production, so speed, accuracy, and stability are important. You will help us: Build ...generation Maintain strong prompt + visual accuracy Produce deterministic results Deliver a clean inference API (REST or gRPC) Performance target: Image generation in about 5–8 seconds on a 4090 GPU. We want someone who can work with us long-term as we continue developing features like LoRAs, presets, and video generation later on. If you have experience with diffusion models, ControlNet, SDXL, LoRA training, or production deployment, we want to ...

...FUNDRAISING INTEGRATION Use: Stripe PayPal Flutterwave (optional for African members) Blockchain wallet support (optional future feature) Features: Transparent donation logs Donor profiles Project progress tracking Exportable financial reports 8. PLATFORM BRANDING FOR DEVELOPERS Colors Gold: #D4AF37 Navy Blue: #021A4E White: #FFFFFF Champagne: #F8F4EF Fonts Headers: Cinzel / Playfair Display Body: Lora / Garamond Logo Usage Full Crest only on official pages Mini Crest on dashboard header Gold-only favicon...

...via BLIP or any kohya_ss-compatible module). Generated captions should be written back as *.txt files next to each image so kohya_ss can pick them up for training. 2. LoRA training Using those freshly captioned images, the next section should fine-tune a Stable Diffusion-compatible model into a LoRA. I want the usual kohya_ss training options exposed (learning rate, resolution, batch size, number of epochs) but filled with sensible defaults. The code should automatically save the resulting *.safetensors LoRA to Drive. 3. Image generation with background control Finally, include an inference cell that loads the trained LoRA, merges it with the base model if needed, and produces exactly five images per run. The prompt must let me pin the subject’s o...

...production-ready design that turns that sketch into Gerber files and a clean bill of materials. Key points you’ll be working with: • The Xiao RP2350 must remain the central MCU footprint. • I need on-board power-supply circuitry so the board can run from a single external DC source without extra regulators. • Space and routing for at least one communication module (for example BLE, Wi-Fi or LoRa) must be included; exact part number can be finalised together, but the footprint and connections have to be there from the start. • No dedicated sensors are required right now—this board is for rapid prototyping. Deliverables I expect: • Schematic in KiCad, Altium, or another mainstream EDA tool we agree on. • Gerber, drill and pick...

I need my existing hardware brought up to date. I will share the current schematics and board files, and from those references I’d like you to: • Refresh the schematic, correcting any outdated footprints or pin mappings and integrating the ESP32, an LTE module, and a LoRa module exactly as shown in my reference design. • Redesign the PCB for a clean, manufacturable layout—four-layer if necessary—with good RF isolation between the LTE and LoRa sections and solid power-supply routing for the ESP32. • Deliver production-ready Gerbers, an up-to-date BOM, and source files (KiCad or Altium are both fine). I’ll review the schematic first, then the routed board. Once the final files pass DRC/ERC and my manufacturing checks, the job is ...

...they retain • the identical clothing style and fit • the original pose and background of the target shot • an exact match of the target person’s facial expression Reliability is key: every result should look like the same outfit was tailored to the new physique without distortion or mismatched wrinkles. I am open to your preferred deep-learning stack (e.g., Stable Diffusion, ControlNet, LoRA fine-tuning, face-landmark-driven warping, etc.) as long as it installs smoothly in a Python environment and consumes the standard GPU instance already available on Runninghub. Deliverables • Source code or notebook with clear, reproducible steps • Requirements file / Docker snippet so the setup works out-of-the-box on Runninghub • Simple C...

...naturally in future replies Vector database (Weaviate, Qdrant, Milvus, etc.) + proper chunking & retrieval strategy Hybrid approach: RAG + optional light fine-tuning (LoRA/QLoRA) on our data to make the style even closer to our real conversations Horizontal scaling ready (Docker/Kubernetes or at least very easy to scale) Webhook/integration docs so our existing dating sites and apps can connect in one line of code Required Experience Already built and deployed real chatbots using open-source LLMs (Llama 3/3.1, Mistral, Mixtral, etc.) Proven experience with RAG on large datasets (1M+ documents) Experience training or fine-tuning (LoRA) models for specific tones/styles Previously created bots with adult/flirty/role-play personalities (this is crucial – regular &...