Hacker Read top | best | new | newcomments | leaders | about | bookmarklet login

Not saying anything about training those models, but I can run the weights of Stable Diffusion without larger problems on a vintage RTX 1080Ti.


sort by: page size:

The models are expensive to train right now, but I suspect in 10 years, anyone with a multi gpu rig could train the equivalent of Stable Diffusion.

for what it's worth, stable diffusion was trained on 32 x 8 x A100 GPUs

Stable Diffusion is an image generation model that's been released to the public at large. If you have a decent GPU, you can run the model yourself. (Even without a decent GPU technically you can still do it, though it's much slower)

Stable Diffusion works for me with a Polaris GPU. Had to compile my own local copy of Tensorflow to use it, but everything runs.

The Stable Diffusion models are very small though. You can probably train one with relatively low investment, e.g. 4x3090 under $20k.

From what I've seen, it's possible to take a version of Stable Diffusion and add your training set on top.

What sort of setup do you need to be able to fine tune Stable Diffusion models? Are there good tutorials out there for fine tuning with cloud or non-cloud GPUs?

Many old consumer gaming GPUs will run an implementation of Stable Diffusion. But this page seems to be about getting use of H100 and A100, such as one might want for running or training decent-sized LLMs.

It might require dedicated hardware. That only really becomes possible when you've proven the idea, but ASICs for cryptomining, TensorFlow, etc are quite real. There's no reason why dedicated hardware for training Stable Diffusion couldn't happen.

A lot of HN has been having fun with stable diffusion. Do we really need 1 x GPU with 10GB of RAM? How do you distribute or "shard" a model you're training? Could we get this running on the raspberry pi clusters we all have? Hook it up to OpenFaaS too.

Stable Diffusion runs really well on M1 GPU.

You can run stable diffusion on a MBP and produce images in under a minute. It's training these models that takes the crazy GPU power - running them is quite reasonable.

With stable diffusion Im making right now an image every 26 seconds at 512x512 with 50 sampling steps with https://github.com/JoePenna/Dreambooth-Stable-Diffusion

The training with a beefy GPU from vast.ai (RTX 3090 with 24vram) and Im generating the images with a GTX 1080 with 4vram, so no need for 6 or even 10 GVram from my testing


Hopefully stability.ai will train this model and release it open-source. It's much easier to train it than Stable Diffusion after all.

Yes, Stable diffusion is an open source AI image generator that runs on your own hardware.

No one uses raw stable diffusion though, there are model mixes for whatever usecase you have.

Stable Diffusion has been trained at 512x512 and doesn't work very well above this. But upscalers are ok and can even run on CPUs.

Stable Diffusion has a smaller text encoder than Dalle 2 and other models (Imagen, Parti, Craiyon) so that it can fit into consumer GPUs. I believe StabilityAI will train models based on a larger text encoder, the text encoder is frozen and does not require training, so scaling the text encoder is quite free. For now this is the biggest bottleneck with Stable Diffusion, the generator is really good and the image quality alone is incredible (managing to outperform Dalle 2 most of the time).

I agree, I've definitely seen way more information about running image synthesis models like Stable Diffusion locally than I have LLMs. It's counterintuitive to me that Stable Diffusion takes less RAM than an LLM, especially considering it still needs the word vectors. Goes to show I know nothing.

I guess it comes down to the requirement of a very high end (or multiple) GPU that makes it impractical for most vs just running it in Colab or something.

Tho there are some efforts:

https://github.com/cocktailpeanut/dalai

next

Legal | privacy