DeepForgeby svx

0/5730 solved

← All interview tracks

Inference · serving engineersQuantization · attention kernels · serving

ML Systems · Inference Engineer

Serving a model in production: quantization and KV-cache accounting, batching and latency budgets, then fused kernels and speculative decoding.

3 phases · 26 problems · 26 unique

Start timed mock5 problems · 4 min each

Your progress

0/26 · 0%

Continue track · Model Size (MB)

1Warm-up

0/8

Model Size (MB)
dl-057Deep Learning
Easy
Int8 Quantization Scale
dl-058Deep Learning
Easy
Quantize Int8
dl-059Deep Learning
Easy
Dequantize Int8
dl-060Deep Learning
Easy
KV Cache Size
dl-075Deep Learning
Medium
Throughput Estimate
dl-101Deep Learning
Easy
Latency Estimate
dl-102Deep Learning
Easy
KV Cache Append Step
dl-186Deep Learning
Easy

2Core

0/10

Per-Channel Quant Scale
dl-077Deep Learning
Medium
Weight-Only Quant Memory
dl-195Deep Learning
Easy
Online Softmax Rescale Step
dl-211Deep Learning
Medium
Dynamic Batch Admission
dl-213Deep Learning
Medium
SmoothQuant Migration Factor
dl-214Deep Learning
Medium
Quantized Inference Latency
dl-301Deep Learning
Medium
Batch-Latency Tradeoff
dl-305Deep Learning
Medium
Throughput Saturation
dl-306Deep Learning
Medium
Latency SLA Budget
dl-307Deep Learning
Medium
KV Cache Eviction Policy
dl-308Deep Learning
Medium

3Hard

0/8

Flash Decode Block Split
dl-224Deep Learning
Hard
Roofline Bound Check
dl-228Deep Learning
Hard
MFU Utilization
dl-229Deep Learning
Hard
Fused Attention SRAM Tiles
dl-272Deep Learning
Hard
Speculative Speedup Estimate
dl-315Deep Learning
Hard
Continuous Batching Slot Simulation
dl-316Deep Learning
Hard
CPU Offload Time
dl-318Deep Learning
Hard
Register Pressure Estimate
dl-320Deep Learning
Hard