MosaicML Diffusion

Closed weights Databricks 1.3B parameters April 2023

No estimate

No hardware requirements for this model

The weights for this model have not been published, so it cannot be downloaded or run on your own hardware at any size. It is reachable only through its provider, and no graphics card changes that.

On record

Full specification

Everything on record for this model. Most of it describes how it was trained rather than how it runs — useful context for judging how much work went into it, and how it compares with models built at a different scale.

Origin

Who built this model, where, and when it was published.

Organisation
Databricks
Organisation type
Industry
Country
United States of America
Published
28 April 2023
Authors
Mihir Patel, Erica Ji Yuen, Cory Stephenson, Landan Seguin

What it does

The problem areas the model was built for. A model can carry several of each.

Domain
Image generation
Task
Image generation, Text-to-image
Approach
Self-supervised learning

Size

How large the model is and how much data it was trained on. Parameters are the figure that decides whether it fits on a given graphics card.

Parameters
1.3B

Manually loaded model with following code snippet: from diffusers import DiffusionPipeline, DPMSolverMultistepScheduler import torch repo_id = "stabilityai/stable-diffusion-2" pipe = DiffusionPipeline.from_pretrained(repo_id, torch_dtype=torch.float16, revision="fp16") n_params_vae = sum(p.numel() for p in pipe.components['vae'].parameters()) n_params_text_encoder = sum(p.numel() for p in pipe.components['text_encoder'].parameters()) n_params_unet = sum(p.numel() for p in pipe.components['une…

Training data
790,000,000 tokens

For the first phase of training, we used all images with resolution >=256x256, amounting to 790 million image-caption samples. For the second phase of training, we only used images with resolution >=512x512, amounting to 300 million image-caption samples. Note that second phase data is a strict subset of first stage.

Epochs
1.38

Training compute

The arithmetic performed to train the model, measured in floating-point operations. It is a measure of what the training run cost, not of how fast the finished model answers you.

Training compute
1.1 × 10²² FLOP

23,835 A100-hours * 3.12e14 FLOP/GPU-sec * 3600 sec/hour * 0.4 (assumed utilization) = 1.07e22 "Both phases of training ran on 128 NVIDIA A100 GPUs. The first training phase was run for 550k iterations in 1.6 days while the second phase was run for 850k iterations in 4.9 days, for a total of 20,051 A100 hours for training. In addition to the training time, we pre-computed the latents for the VAE and CLIP model to reduce training time and cost when making multiple passes over the dataset. Pre-co…

How it was established
Hardware

The training run

What it physically took to train: which chips, how many, for how long, and what that drew from the wall.

Training hardware
NVIDIA A100
Chips used
128
Wall-clock time
186 hours (7.8 days)

"Both phases of training ran on 128 NVIDIA A100 GPUs. The first training phase was run for 550k iterations in 1.6 days while the second phase was run for 850k iterations in 4.9 days, for a total of 20,051 A100 hours for training. In addition to the training time, we pre-computed the latents for the VAE and CLIP model to reduce training time and cost when making multiple passes over the dataset. Pre-computing the latents required an additional 3,784 A100 hours, resulting in 23,835 A100 hours in t…

Power draw
102.0 kW

Availability

Whether you can obtain the model and run it on your own hardware, which is what decides if any of the graphics-card figures on this page apply.

Weights
Closed — provider access only
Model access
Unreleased

How it is classified

Labels the source dataset applies when tracking notable models, and how confident it is in the entry.

Record confidence
Confident

Sources

Where this record came from and when it was last checked.

Reference
Training Stable Diffusion from Scratch for <$50k with MosaicML
Last updated
28 November 2025

What the numbers mean

About this model

MosaicML Diffusion was published by Databricks, in the country recorded as United States of America, during April 2023. The publishing organisation is categorised as industry.

It works in the domain of Image generation, and is recorded as performing the task of image generation, Text-to-image.

Because the weights are not available, none of the hardware figures elsewhere on this site apply to it.

What went into building it

Training it took a computation budget of roughly 1.1 × 10²² FLOP, on hardware recorded as NVIDIA A100. That figure measures what producing the model cost, and has no bearing on how fast it answers.

The training set ran to roughly 790,000,000 tokens of text.

Answers

MosaicML Diffusion — common questions

01

MosaicML Diffusion— when was it released?

It was published in April 2023. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.

02

MosaicML Diffusion— what is it used for?

It works in the domain of Image generation, and is recorded as handling the task of image generation, Text-to-image. A model can carry several of each, so these are the areas it was built for rather than a limit on what it will attempt.

03

MosaicML Diffusion— how much compute was used to train it?

Training consumed around 1.1 × 10²² FLOP, on hardware recorded as NVIDIA A100. That measures what producing the model cost and says nothing about how quickly it answers once trained — inference speed comes from memory bandwidth, not from the training budget.

04

MosaicML Diffusion— what GPU do I need to run it?

None. This is a closed model — its weights were never published, so it cannot be downloaded or run on your own hardware at any price. It is reachable only through its provider.

05

MosaicML Diffusion— is it open source?

No. Its weights have not been published, so it exists only as a service controlled by its owner.

06

MosaicML Diffusion— how many parameters does it have?

It has a parameter count of 1.3B. Manually loaded model with following code snippet: from diffusers import DiffusionPipeline, DPMSolverMultistepScheduler import torch repo_id = "stabilityai/stable-diffusion-2" pipe = DiffusionPipeline.from_pretrained(repo_id, torch_dtype=torch.float16, revision="fp16") n_params_vae = sum(p.numel() for p in pipe.components['vae'].parameters()) n_params_text_encoder = sum(p.numel() for p in pipe.components['text_encoder'].parameters()) n_params_unet = sum(p.numel() for p in pipe.components['unet'].parameters()) n_params = n_params_vae + n_params_text_encoder + n_params_unet print(f"Total number of parameters: {n_params}"). That figure is the total, and it is what decides how much memory the model needs — roughly half a gigabyte per billion at the compression most people use.

07

MosaicML Diffusion— who created it?

It was published by Databricks, based in United States of America, an organisation categorised as industry.

Source

Original publication

Record last updated 28 November 2025

The other direction

Looking at it from the other side?

This page starts from the model. If you already own a card and want to know everything it will run, start from the hardware instead.