AI Automation
Courses
5 course options · Compare the learning

LLM Inference and Serving Optimization Courses

Compare engineering curricula for reducing inference latency or memory and serving workloads through runtime mechanisms.

Illustrative workflow

A concrete way to compare the learning

Model/workload → quantize/cache/batch/parallelize → serve → compare latency, memory and quality.

Check it: Use the named comparison dimensions to check the result; a demo, learner exercise and assessed project are distinct.

01 / Start withModel/workload02 / Comparequantize/cache/batch/parallelize →serve03 / Useful outputcompare latency, memory andquality.
One task, three points to compare.

This is a decision aid, not a promised assignment in every listed course.

Compare the course work and setup.

A demonstration, a learner exercise and individually reviewed work are different learning experiences. Check the practice and feedback each option actually describes.

Course and providerPracticeSetupTuition and format
Fast & Efficient LLM Inference with vLLMDeepLearning.AIA full-precision Qwen model is compressed with LLM Compressor; size/perplexity checks guide deployment through vLLM; OpenAI-compatible requests and metrics lead to GuideLLM load…Python and basic token/inference/GPU concepts are recommended.Price: ask the provider. Price basis: Own page advertises Enroll Now and additional Pro quizzes/projects/accomplishment. Base lesson-access price and Pro…Online video lessons with advertised examples/resources; the page lists 3 code examples. Live cohort, mentoring and project access are not stated.
Efficiently Serving LLMsDeepLearning.AIToken generation, KV caching, batching and quantization are implemented and benchmarked; LoRA and multi-LoRA LoRAX serving demonstrate throughput/latency tradeoffs.Intermediate Python knowledge required by the public description.Free learning option. Price basis: Own page advertises free course access for a limited time during platform beta; not perpetual free access or a Pro…Online video lessons with advertised examples/resources; the page lists 6 code examples. Live cohort, mentoring and project access are not stated.
Efficient Inference with SGLang: Text and Image GenerationDeepLearning.AIAttention and KV cache are implemented before radix-tree shared-prefix caching across requests; caching and parallelism are applied to text and diffusion image generation, with…Python and basic language-model knowledge are recommended.Price: ask the provider. Price basis: Own page advertises Enroll Now and additional Pro quizzes/projects/accomplishment. Base lesson-access price and Pro…Online video lessons with advertised examples/resources; the page lists 3 code examples. Live cohort, mentoring and project access are not stated.
Quantization in DepthDeepLearning.AIA custom PyTorch quantizer implements symmetric/asymmetric and tensor/channel/group schemes; quantization errors, size/performance and low-bit packing are evaluated.Builds on Quantization Fundamentals; prior relevant foundation/PyTorch context expected.Free learning option. Price basis: Own page advertises free course access for a limited time during platform beta; not perpetual free access or a Pro…Online video lessons with advertised examples/resources; the page lists 13 code examples. Live cohort, mentoring and project access are not stated.
Quantization Fundamentals with Hugging FaceDeepLearning.AILinear quantization, downcasting and Quanto are applied through Transformers to open language/multimodal models, with memory/quality effects examined before the advanced quantizer…Basic machine learning and some PyTorch expected.Free learning option. Price basis: Own page advertises free course access for a limited time during platform beta; not perpetual free access or a Pro…Online video lessons with advertised examples/resources; the page lists 3 code examples. Live cohort, mentoring and project access are not stated.

Course options

Look beyond the title. Compare the project work, required accounts, software costs and type of help.

Browse all courses
1–5 of 5 courses
AI engineering

DeepLearning.AI

Fast & Efficient LLM Inference with vLLM

A full-precision Qwen model is compressed with LLM Compressor; size/perplexity checks guide deployment through vLLM; OpenAI-compatible requests and metrics lead to GuideLLM load tests and lm-eval quality…

Tuition
Price: ask the provider. Price basis: Own page advertises Enroll Now and additional Pro quizzes/projects/accomplishment. Base lesson-access price and Pro…
Format
Online video lessons with advertised examples/resources; the page lists 3 code examples. Live cohort, mentoring and project access are not stated.
Practice
A full-precision Qwen model is compressed with LLM Compressor; size/perplexity checks guide deployment through vLLM; OpenAI-compatible requests and metrics lead to GuideLLM load…
Setup
Python and basic token/inference/GPU concepts are recommended.
Course details ↗Provider access conditions apply
AI engineering

DeepLearning.AI

Efficiently Serving LLMs

Token generation, KV caching, batching and quantization are implemented and benchmarked; LoRA and multi-LoRA LoRAX serving demonstrate throughput/latency tradeoffs.

Tuition
Free learning option. Price basis: Own page advertises free course access for a limited time during platform beta; not perpetual free access or a Pro…
Format
Online video lessons with advertised examples/resources; the page lists 6 code examples. Live cohort, mentoring and project access are not stated.
Practice
Token generation, KV caching, batching and quantization are implemented and benchmarked; LoRA and multi-LoRA LoRAX serving demonstrate throughput/latency tradeoffs.
Setup
Intermediate Python knowledge required by the public description.
Course details ↗Provider access conditions apply
AI engineering

DeepLearning.AI

Efficient Inference with SGLang: Text and Image Generation

Attention and KV cache are implemented before radix-tree shared-prefix caching across requests; caching and parallelism are applied to text and diffusion image generation, with performance tradeoffs examined.

Tuition
Price: ask the provider. Price basis: Own page advertises Enroll Now and additional Pro quizzes/projects/accomplishment. Base lesson-access price and Pro…
Format
Online video lessons with advertised examples/resources; the page lists 3 code examples. Live cohort, mentoring and project access are not stated.
Practice
Attention and KV cache are implemented before radix-tree shared-prefix caching across requests; caching and parallelism are applied to text and diffusion image generation, with…
Setup
Python and basic language-model knowledge are recommended.
Course details ↗Provider access conditions apply
AI engineering

DeepLearning.AI

Quantization in Depth

A custom PyTorch quantizer implements symmetric/asymmetric and tensor/channel/group schemes; quantization errors, size/performance and low-bit packing are evaluated.

Tuition
Free learning option. Price basis: Own page advertises free course access for a limited time during platform beta; not perpetual free access or a Pro…
Format
Online video lessons with advertised examples/resources; the page lists 13 code examples. Live cohort, mentoring and project access are not stated.
Practice
A custom PyTorch quantizer implements symmetric/asymmetric and tensor/channel/group schemes; quantization errors, size/performance and low-bit packing are evaluated.
Setup
Builds on Quantization Fundamentals; prior relevant foundation/PyTorch context expected.
Course details ↗Provider access conditions apply
AI engineering

DeepLearning.AI

Quantization Fundamentals with Hugging Face

Linear quantization, downcasting and Quanto are applied through Transformers to open language/multimodal models, with memory/quality effects examined before the advanced quantizer follow-up.

Tuition
Free learning option. Price basis: Own page advertises free course access for a limited time during platform beta; not perpetual free access or a Pro…
Format
Online video lessons with advertised examples/resources; the page lists 3 code examples. Live cohort, mentoring and project access are not stated.
Practice
Linear quantization, downcasting and Quanto are applied through Transformers to open language/multimodal models, with memory/quality effects examined before the advanced quantizer…
Setup
Basic machine learning and some PyTorch expected.
Course details ↗Provider access conditions apply

Related learning decisions