From 249fadfb768eb4c49f4f42dcad0b91d9beeadce7 Mon Sep 17 00:00:00 2001 From: OleehyO Date: Wed, 8 Jan 2025 01:39:37 +0000 Subject: [PATCH] docs: add hardware requirements for model training Add a table in README files showing hardware requirements for training different CogVideoX models, including: - Memory requirements for each model variant - Supported training types (LoRA) - Training resolutions - Mixed precision settings Updated in all language versions (EN/ZH/JA). --- finetune/README.md | 11 +++++++++++ finetune/README_ja.md | 11 +++++++++++ finetune/README_zh.md | 14 ++++++++++++++ 3 files changed, 36 insertions(+) diff --git a/finetune/README.md b/finetune/README.md index 1667429..156d5c8 100644 --- a/finetune/README.md +++ b/finetune/README.md @@ -6,6 +6,17 @@ If you're looking for the fine-tuning instructions for the SAT version, please check [here](../sat/README_zh.md). The dataset format for this version differs from the one used here. +## Hardware Requirements + +| Model | Training Type | Mixed Precision | Training Resolution (frames x height x width) | Hardware Requirements | +|---------------------|-----------------|----------------|---------------------------------------------|------------------------| +| cogvideox-t2v-2b | lora (rank128) | fp16 | 49x480x720 | 16GB VRAM (NVIDIA 4080) | +| cogvideox-t2v-5b | lora (rank128) | bf16 | 49x480x720 | 24GB VRAM (NVIDIA 4090) | +| cogvideox-i2v-5b | lora (rank128) | bf16 | 49x480x720 | 24GB VRAM (NVIDIA 4090) | +| cogvideox1.5-t2v-5b | lora (rank128) | bf16 | 81x768x1360 | 35GB VRAM (NVIDIA A100) | +| cogvideox1.5-i2v-5b | lora (rank128) | bf16 | 81x768x1360 | 35GB VRAM (NVIDIA A100) | + + ## Install Dependencies Since the relevant code has not yet been merged into the official `diffusers` release, you need to fine-tune based on the diffusers branch. Follow the steps below to install the dependencies: diff --git a/finetune/README_ja.md b/finetune/README_ja.md index 5730974..8df7c4b 100644 --- a/finetune/README_ja.md +++ b/finetune/README_ja.md @@ -6,6 +6,17 @@ SATバージョンのファインチューニング手順については、[こちら](../sat/README_zh.md)をご確認ください。このバージョンのデータセットフォーマットは、こちらのバージョンとは異なります。 +## ハードウェア要件 + +| モデル | トレーニングタイプ | 混合精度学習 | トレーニング解像度(フレーム数x高さx幅) | ハードウェア要件 | +|----------------------|-----------------|------------|----------------------------------|----------------| +| cogvideox-t2v-2b | lora (rank128) | fp16 | 49x480x720 | 16GB VRAM (NVIDIA 4080) | +| cogvideox-t2v-5b | lora (rank128) | bf16 | 49x480x720 | 24GB VRAM (NVIDIA 4090) | +| cogvideox-i2v-5b | lora (rank128) | bf16 | 49x480x720 | 24GB VRAM (NVIDIA 4090) | +| cogvideox1.5-t2v-5b | lora (rank128) | bf16 | 81x768x1360 | 35GB VRAM (NVIDIA A100) | +| cogvideox1.5-i2v-5b | lora (rank128) | bf16 | 81x768x1360 | 35GB VRAM (NVIDIA A100) | + + ## 依存関係のインストール 関連するコードがまだ `diffusers` の公式リリースに統合されていないため、`diffusers` ブランチを基にファインチューニングを行う必要があります。以下の手順に従って依存関係をインストールしてください： diff --git a/finetune/README_zh.md b/finetune/README_zh.md index e40ef5c..d72a935 100644 --- a/finetune/README_zh.md +++ b/finetune/README_zh.md @@ -6,6 +6,20 @@ 如果您想查看SAT版本微调，请查看[这里](../sat/README_zh.md)。其数据集格式与本版本不同。 +## 硬件要求 + +| 模型 | 训练类型 | 混合训练精度 | 训练分辨率(帧数x高x宽) | 硬件要求 | +|----------------------|----------------|------------|----------------------|-----------------------| +| cogvideox-t2v-2b | lora (rank128) | fp16 | 49x480x720 | 16G显存 (NVIDIA 4080) | +| cogvideox-t2v-5b | lora (rank128) | bf16 | 49x480x720 | 24G显存 (NVIDIA 4090) | +| cogvideox-i2v-5b | lora (rank128) | bf16 | 49x480x720 | 24G显存 (NVIDIA 4090) | +| cogvideox1.5-t2v-5b | lora (rank128) | bf16 | 81x768x1360 | 35G显存 (NVIDIA A100) | +| cogvideox1.5-i2v-5b | lora (rank128) | bf16 | 81x768x1360 | 35G显存 (NVIDIA A100) | + + ## 安装依赖 由于相关代码还没有被合并到diffusers发行版，你需要基于diffusers分支进行微调。请按照以下步骤安装依赖：