Built for US teams
Payment by wire transferEnhanced confidentialityPerfect for companies with 50 to 500 employeesScheduling to fit any time zone
← Back

Training TensorRT-LLM - Optimizing LLM Inference in Production

Ref: YRF173
10 people max.
From $5,880 HT / per person
Pay in 3 installments · On-site on request · +$540 with certification exam
4 days
Remote

Share in 2 clicks

EquansEDFPhotowattAptarArcelorMittalUbisoftINSEECLa PlateformeCESIEFREIEPSIINGETISMy Digital SchoolYnovEquansEDFPhotowattAptarArcelorMittalUbisoftINSEECLa PlateformeCESIEFREIEPSIINGETISMy Digital SchoolYnov

Learning objectives

  • Master TensorRT-LLM to accelerate LLM inference in professional production environments
  • Develop optimized TensorRT engines for certifiable enterprise competencies
  • Implement quantization techniques and layer fusion with TensorRT-LLM
  • Design scalable and high-performance inference pipelines for AI applications
  • Optimize GPU performance with custom kernels and advanced profiling
  • Deploy TensorRT-LLM solutions integrated with Triton Inference Server
  • Evaluate and tune models to minimize latency and costs in enterprise environments

The Learni story

Founded by engineers and learning experts, Learni's mission is to make high-impact tech training accessible to teams everywhere. We work remotely with organizations across the US and Canada, in your time zone, to help teams upskill fast.

Don't let this gap widen

Why this program matters

  • Without this upskilling, your team accumulates a technological gap that translates directly into productivity loss.

  • Organizations that don't train their talents on key topics see their competitiveness drop.

  • Every quarter without training is a gap widening with competitors who invest.

  • The cost of inaction quickly exceeds that of well-targeted training.

Imed BEN AMOR
Imed BEN AMOR

Learni trainer · Development expert

73%productivity gap
×3cost of inaction

Program

Module 1TensorRT-LLM Fundamentals: Installation and Engine Building (NVIDIA Docker, HuggingFace)

Complete installation of the TensorRT-LLM environment on NVIDIA GPU, hands-on with Docker and CUDA Toolkit tools to create a first engine from a HuggingFace model like Llama-2, practical exercises on HF to TensorRT conversion, verification of initial performance with integrated benchmarks, generation of latency and throughput reports, setup of an ongoing enterprise LLM project to measure immediate gains.

Module 2Advanced TensorRT-LLM Optimization: Quantization and Fusion (FP8, GEMM plugins)

Exploration of INT4/INT8/FP8 quantization techniques in TensorRT-LLM, automatic fusion of attention and MLP layers to reduce GPU memory, development of custom GEMM plugins to accelerate matrix calculations, practical cases on models like Mistral with hyperparameter tuning, profiling with Nsight Systems to identify bottlenecks, production of an optimized engine delivering up to 3x speedup, documentation of technical choices for production replication.

Module 3TensorRT-LLM Deployment: Integration and Scaling (Triton Server, KV cache)

Integration of TensorRT-LLM with NVIDIA Triton Inference Server for multi-model deployments, management of KV cache for long-contextual streaming inferences, configuration of dynamic batching and tensor parallelism on GPU clusters, exercises on horizontal scaling with Kubernetes, load testing with Locust to simulate 1000+ requests/second, resolution of real enterprise error cases, deployment of a secure production-ready REST API service.

Module 4TensorRT-LLM Performance Tuning: Profiling and Real Cases (Nsight, multi-GPU)

In-depth performance analysis with NVIDIA Nsight Compute and DCGM to tune TensorRT-LLM kernels, multi-GPU optimization with tensor and pipeline parallelism on 70B parameter LLMs, resolution of challenges like Out-Of-Memory via intelligent paging and swapping, workshops on concrete enterprise cases like scalable chatbots, finalization of the ongoing project with live Q&A, evaluation of final gains (latency <50ms, 70% cost reduction), delivery of reusable certifiable templates.

Evaluation method

  • Technical multiple-choice quiz on TensorRT-LLM and its optimizations
  • Practical evaluation through development of a custom engine
  • Presentation of the ongoing project with performance measurements

Learning method

  • Courses led by an active NVIDIA expert trainer
  • Practical exercises on real enterprise LLM cases
  • Progressive ongoing project over 4 days
  • Complete course materials and source codes provided

Methods, materials and delivery

The Training TensorRT-LLM - Optimizing LLM Inference in Production program is delivered onsite or remote (blended-learning, virtual classroom, remote presence). At Learni, an industry-certified training organization, every program is built to maximize skills acquisition regardless of the chosen format.

The trainer alternates between demonstrative, interrogative and active methods (through hands-on labs and/or scenarios). This pedagogical approach guarantees concrete learning that's immediately applicable at work.

Equipment required

For the smooth delivery of the Training TensorRT-LLM - Optimizing LLM Inference in Production program, the following equipment is required:

  • Mac or PC computers, high-speed fiber internet, whiteboard or flipchart, projector or interactive touch screen (for remote sessions)
  • Training environments installed on workstations or accessible online
  • Course materials, hands-on exercises and complementary resources
  • Post-training access to materials and educational resources

For intra-company training on a site outside Learni, the client commits to providing all required teaching materials (computers, internet, etc.) for the smooth delivery of the program in line with the prerequisites in the communicated program.

* contact us for remote delivery feasibility** ratio varies depending on the program

Skills assessment methods

Assessment of skills acquired during the Training TensorRT-LLM - Optimizing LLM Inference in Production program is performed through:

  • During training: case studies, hands-on labs and professional scenarios
  • End of training: self-assessment questionnaire and skills evaluation by the trainer
  • After training: completion certificate detailing acquired skills

Program accessibility

Learni is committed to making its programs accessible. All our programs are accessible to people with disabilities. Our teams are available to adapt the pedagogical methods to your specific needs. Please contact us for any adjustment request.

Enrollment terms and lead times

Registration is possible up to 48 business hours before the start of training. All our programs are built for corporate L&D budgets and delivered onsite or remotely.

Our method

Training quality, guaranteed at every step

Before, during, after: we frame the brief, introduce the trainer, tailor the content and measure impact. You stay in control from kickoff to wrap-up.

Step 1

Rigorous trainer selection

Each trainer is validated on three criteria: hands-on field expertise, proven pedagogy and alignment with your industry.

  • Triple validation: technical, pedagogical, sectoral.
  • Minimum rating 4.8/5 over the last 12 sessions.
Step 2

You meet the trainer beforehand

30-minute video call between you and the selected trainer to validate the fit, adjust content and clear any final doubts.

  • Live briefing on goals and team context.
  • Veto right — we swap the trainer for free if needed.
Step 3

Content tailored to your context

No recycled slides. The syllabus is reworked from your real cases: tools, constraints, vocabulary, ongoing projects.

  • Hands-on cases drawn from your stack and projects.
  • Program co-written then validated by your team.
Step 4

Continuous quality follow-up

Live evaluations, 30/90/180-day check-ins and a consolidation plan. If the impact misses the mark, we rework it.

  • NPS, knowledge quizzes and skills self-assessment.
  • Satisfaction guarantee: fully satisfied or free rework.

A simple promise: you don't pay to discover the trainer on day one. Everything is validated upfront, by you.

FAQ

Frequently asked questions

How much does the Training TensorRT-LLM - Optimizing LLM Inference in Production training cost?+
The individual price is $5,880 (USD). A detailed quote is sent within one business day.
How long is the Training TensorRT-LLM - Optimizing LLM Inference in Production training?+
The training lasts 4 journées, available live online (US time zones) or on-site at your offices.
How is this training paid for?+
Most US teams pay directly through their company (L&D or training budget). We invoice in US dollars and accept bank transfer (ACH/wire) or card, with volume pricing for teams. A purchase order is welcome.
Are there any prerequisites?+
Expertise in PyTorch/TensorFlow, CUDA programming, and deployment of LLM models such as Llama or GPT
Is a certificate delivered at the end?+
Yes. A Learni completion certificate is issued, along with the individual evaluation report.
Does Learni provide the equipment?+
No. A computer and stable internet connection are required for the participant. Learni provides the educational platform, the trainer and all course materials.
On-site & remote

This training across cities

Available on-site and remotely. Pick your city to see the local training center.

Your professional training, anywhere

Let's build
your next
program.

30 minutes with a learning advisor. No commitment. No sales pitch dressed up as a demo.

Reply within 24h · Vetted experts · USD invoicing · W-9 available