From f044191cfbbe14fa9f44fedc9b23b050af780c51 Mon Sep 17 00:00:00 2001 From: Murat Raimbekov Date: Fri, 24 Apr 2026 03:00:49 +0600 Subject: [PATCH] Remove drifting GPU heuristic table from platform quickstart (#24303) --- docs/en/platform/quickstart.md | 16 ++-------------- 1 file changed, 2 insertions(+), 14 deletions(-) diff --git a/docs/en/platform/quickstart.md b/docs/en/platform/quickstart.md index 6afd18b1..f44b1cd5 100644 --- a/docs/en/platform/quickstart.md +++ b/docs/en/platform/quickstart.md @@ -237,21 +237,9 @@ From your project, click `Train Model` to start cloud training. ### Training Configuration 1. **Select Dataset**: Choose from your uploaded datasets (only datasets with a [`train` split](data/datasets.md#filter-by-split) are shown) -2. **Choose Model**: Select a base model — official Ultralytics models or your own trained models +2. **Choose Model**: Select a base model - official Ultralytics models or your own trained models 3. **Set Epochs**: Number of training iterations (default: 100) -4. **Select GPU**: Choose compute resources based on your budget and model size - -| Model | Size | Speed | Accuracy | Recommended GPU | -| ------- | ----------- | -------- | -------- | ----------------------------------------------------------------------- | -| YOLO26n | Nano | Fastest | Good | RTX PRO 6000 (96 GB) or RTX 4090 (24 GB) | -| YOLO26s | Small | Fast | Better | RTX PRO 6000 (96 GB) | -| YOLO26m | Medium | Moderate | High | RTX PRO 6000 (96 GB) | -| YOLO26l | Large | Slower | Higher | RTX PRO 6000 (96 GB) or A100 SXM (80 GB) | -| YOLO26x | Extra Large | Slowest | Best | H100 SXM (80 GB); H200 (141 GB) or B200 (180 GB) with Pro or Enterprise | - -!!! info "GPU Selection" - - GPUs range from $0.24/hr (RTX 2000 Ada, 16 GB) to $4.99/hr (B200, 180 GB). The default GPU is **RTX PRO 6000** (96 GB Blackwell, $1.69/hr) — a great balance of memory and performance. 20 GPUs are available on all plans; H200 and B200 require [Pro or Enterprise](account/billing.md#plans). See the full [GPU pricing table](index.md#what-gpu-options-are-available-for-cloud-training). +4. **Select GPU**: Choose compute resources based on your budget and model size. The default is **RTX PRO 6000** (96 GB Blackwell, $1.69/hr), which handles every YOLO26 variant. See the full [GPU pricing table](index.md#what-gpu-options-are-available-for-cloud-training) or the [Cloud Training GPU step](train/cloud-training.md#step-5-select-gpu-cloud-tab) for the complete list and tier gating. !!! warning "Credit Balance Required"