Make videos on the computer you already own. FreeVideo runs MiniMax H3 in as little as 8 GB of VRAM and 16 GB of RAM, and adapts its acceleration path to your hardware.
Add a community results page (#113)
Bowen Xue3559d0f6ca11
diff --git a/README.md b/README.mdindex 71c99b8..7e9bb41 100644--- a/README.md+++ b/README.md@@ -94,6 +94,7 @@ Generate a video from a prompt file:### More details- [FreeVideo Adaptive Execution Planner](docs/execution-planning.md)+- [Community results](docs/community-results.md)### Supportdiff --git a/README.zh-CN.md b/README.zh-CN.mdindex c79b817..44c1a47 100644--- a/README.zh-CN.md+++ b/README.zh-CN.md@@ -94,6 +94,7 @@ git clone https://github.com/FlashML-org/FreeVideo.git && cd FreeVideo### 更多文档- [FreeVideo Adaptive Execution Planner](docs/zh-CN/execution-planning.md)+- [社区实测](docs/zh-CN/community-results.md)### 问题反馈diff --git a/docs/community-results.md b/docs/community-results.mdnew file mode 100644index 0000000..4768d44--- /dev/null+++ b/docs/community-results.md@@ -0,0 +1,25 @@+# Community results++Thanks to the community members who sent diagnostic reports. FreeVideo already runs on everything from 8 GB laptops to professional workstations; the table below summarizes the measurements and optimizations from these reports:++| RTX GPU | VRAM | RAM | Resolution | Length | Quality | Version | Time | Optimized in | After |+|---|---:|---:|---:|---:|:---:|:---:|---:|:---:|---:|+| 4060 Ti | 16 GB | 32 GB | 1920×1088 | 15 s | Light | 0.3.0 | 3306 s | 0.3.6 | 2730 s (−580 s) |+| 5060 | 8 GB | 16 GB | 1344×768 | 15 s | Light | 0.3.1 | 1367 s | 0.3.6 | 1160 s (−210 s) |+| 3070 Ti | 8 GB | 96 GB | 1344×768 | 10 s | Light | 0.3.1 | 749 s | 0.3.6 | 690 s (−60 s) |+| 4060 | 8 GB | 32 GB | 768×1344 | 15 s | Light | 0.3.4 | 1592 s | 0.3.10 | 1450 s (−140 s) |+| 5070 Laptop | 8 GB | 32 GB | 1344×768 | 5 s | Light | 0.3.4 | 555 s | 0.3.10 | 485 s (−70 s) |+| PRO 5000 | 72 GB | 128 GB | 768×1344 | 10 s | High | 0.3.4 | 552 s | 0.3.4 | Already optimal |+| 4070 Laptop | 8 GB | 32 GB | 1344×768 | 8 s | Light | 0.3.5 | 1031 s | 0.3.10 | 870 s (−160 s) |+| 3070 Laptop | 8 GB | 32 GB | 1344×768 | 5 s | Light | 0.3.8 | 3821 s | 0.3.8 | Thermal‑limited¹ |+| 5080 Laptop | 16 GB | 32 GB | 1184×672 | 10 s | Light | 0.3.8 | 416 s | 0.3.8 | Already optimal |+| 5090 | 32 GB | 32 GB | 1344×768 | 6.6 s | Max | 0.3.8 | 336 s | 0.3.8 | No report data |+| 5070 Ti | 16 GB | 48 GB | 576×928 | 12.3 s | Medium | 0.3.8 | 472 s | int8 model² | 420 s (−50 s) |+| A5000 Laptop | 16 GB | 128 GB | 1344×768 | 10 s | Light | 0.3.8 | 1108 s | 0.3.8 | Thermal‑limited¹ |+| 4060 Ti | 16 GB | 32 GB | 1344×768 | 15 s | Max | 0.3.8 | 9829 s | 0.3.10 | 2890 s (−6940 s) |++- Time: end to end, from submission until the video is saved.+- After: the estimated time, derived from the report's stage times and from complete planned runs on our test machines (RTX 3090 on Windows, RTX PRO 6000 on Linux) under the reporting machine's VRAM and RAM limits, not from reruns on that machine.+- Already optimal: weight placement and transfers on this machine are already the fastest they can be; the time depends on the GPU's compute and the chosen quality level.+- ¹ Thermal-limited: the GPU spent most of the run in thermal slowdown (3070 Laptop: 229 MHz on average at 89 °C; A5000 Laptop: 882 MHz on average), so cooling limited the time.+- ² This machine still runs the FP8 model; the estimate is for the int8 model, a one-click upgrade in the launcher since 0.3.0.diff --git a/docs/zh-CN/community-results.md b/docs/zh-CN/community-results.mdnew file mode 100644index 0000000..93704d9--- /dev/null+++ b/docs/zh-CN/community-results.md@@ -0,0 +1,25 @@+# 社区实测++感谢提交诊断报告的社区用户。FreeVideo 已经在从 8 GB 显存的笔记本到专业工作站的各类机器上运行,下表汇总了这些报告的实测及优化:++| RTX 显卡 | 显存 | 内存 | 分辨率 | 时长 | 档位 | 版本 | 用时 | 优化版本 | 优化后 |+|---|---:|---:|---:|---:|:---:|:---:|---:|:---:|---:|+| 4060 Ti | 16 GB | 32 GB | 1920×1088 | 15 s | 轻量 | 0.3.0 | 3306 s | 0.3.6 | 2730 s(−580 s) |+| 5060 | 8 GB | 16 GB | 1344×768 | 15 s | 轻量 | 0.3.1 | 1367 s | 0.3.6 | 1160 s(−210 s) |+| 3070 Ti | 8 GB | 96 GB | 1344×768 | 10 s | 轻量 | 0.3.1 | 749 s | 0.3.6 | 690 s(−60 s) |+| 4060 | 8 GB | 32 GB | 768×1344 | 15 s | 轻量 | 0.3.4 | 1592 s | 0.3.10 | 1450 s(−140 s) |+| 5070 笔记本 | 8 GB | 32 GB | 1344×768 | 5 s | 轻量 | 0.3.4 | 555 s | 0.3.10 | 485 s(−70 s) |+| PRO 5000 | 72 GB | 128 GB | 768×1344 | 10 s | 精细 | 0.3.4 | 552 s | 0.3.4 | 已是最优 |+| 4070 笔记本 | 8 GB | 32 GB | 1344×768 | 8 s | 轻量 | 0.3.5 | 1031 s | 0.3.10 | 870 s(−160 s) |+| 3070 笔记本 | 8 GB | 32 GB | 1344×768 | 5 s | 轻量 | 0.3.8 | 3821 s | 0.3.8 | 散热受限¹ |+| 5080 笔记本 | 16 GB | 32 GB | 1184×672 | 10 s | 轻量 | 0.3.8 | 416 s | 0.3.8 | 已是最优 |+| 5090 | 32 GB | 32 GB | 1344×768 | 6.6 s | 极致 | 0.3.8 | 336 s | 0.3.8 | 无诊断数据 |+| 5070 Ti | 16 GB | 48 GB | 576×928 | 12.3 s | 标准 | 0.3.8 | 472 s | int8 模型² | 420 s(−50 s) |+| A5000 笔记本 | 16 GB | 128 GB | 1344×768 | 10 s | 轻量 | 0.3.8 | 1108 s | 0.3.8 | 散热受限¹ |+| 4060 Ti | 16 GB | 32 GB | 1344×768 | 15 s | 极致 | 0.3.8 | 9829 s | 0.3.10 | 2890 s(−6940 s) |++- 用时:从提交到视频保存完成的端到端用时。+- 优化后:预估用时,根据报告中的分段用时,结合测试机(Windows RTX 3090、Linux RTX PRO 6000)在报告机器的显存和内存上限下运行完整规划的实测推算,不是在报告机器上重跑。+- 已是最优:这台机器的权重放置和数据传输已经是最快的方式,用时取决于显卡算力和所选档位。+- ¹ 散热受限:显卡在生成过程中大部分时间处于温控降频(3070 笔记本平均 229 MHz、89 °C;A5000 笔记本平均 882 MHz),用时主要受散热限制。+- ² 这台机器仍在使用 FP8 模型;预估为在启动器中一键升级到 int8 模型(0.3.0 起提供)后的用时。