Ostris's banner

Ostris

@ostrisai • 15,324 subscribers

AI / ML researcher and developer. Creator of AI Toolkit - https://t.co/Thqof0Gxpj Support my work - https://t.co/Isg2EXrP7s

Shorts

I have playing with Black Forest Labs Black Forest Labs FLUX 3 for a few days now. (Thank you for the early access!) And I played with it... a lot. Some may say, too much. 😅 Here are some of my favorite gens. Needless to say, I love this model. Audio on! More in 🧵

I have playing with Black Forest Labs Black Forest Labs FLUX 3 for a few days now. (Thank you for the early access!) And I played with it... a lot. Some may say, too much. 😅 Here are some of my favorite gens. Needless to say, I love this model. Audio on! More in 🧵

32,042 views

I have been tinkering with making a pixel space video model using Wan 2.1 1.3B as the global blocks with modified hourglass transformer blocks on both ends. Just pretraining the hourglass for now. The global blocks have the equivalent of 16 spatial and 8 temporal compression.

I have been tinkering with making a pixel space video model using Wan 2.1 1.3B as the global blocks with modified hourglass transformer blocks on both ends. Just pretraining the hourglass for now. The global blocks have the equivalent of 16 spatial and 8 temporal compression.

25,608 views

Training Wan 2.1-1.3B to use Qwen3-VL-2B text encoder. Doing 33% text only, 33% VL only, and 33% both. Just pretraining the two text input linear layers for now. It is amazing how quickly a model can adapt to a different text encoder. This is 25,750 steps, BS of 10.

Training Wan 2.1-1.3B to use Qwen3-VL-2B text encoder. Doing 33% text only, 33% VL only, and 33% both. Just pretraining the two text input linear layers for now. It is amazing how quickly a model can adapt to a different text encoder. This is 25,750 steps, BS of 10.

26,778 views

A lot of progress has been made on my little pixelspace video model. I am going to need a better dataset. Any suggestions?

A lot of progress has been made on my little pixelspace video model. I am going to need a better dataset. Any suggestions?

13,575 views

Videos

Anya Rossi

sweetdream.ai

SweetDream.ai•Sponsored•Livecam

Watch Anya Live

Anya is streaming live right now! Join her private show and enjoy exclusive content.

Exclusive private shows

1.2k viewers online

Private Show

Join now for exclusive access

Free preview available • Premium content

I trained this LTX LTX 2.3 LoRA of George Costanza at home on my 5090 in about a day with AI Toolkit. I generated this 30 second video with ComfyUI on my 5090 in 6 minutes. Open source is, always has been, and always will be, the future of generative AI. (SOUND ON)

I trained this LTX LTX 2.3 LoRA of George Costanza at home on my 5090 in about a day with AI Toolkit. I generated this 30 second video with ComfyUI on my 5090 in 6 minutes. Open source is, always has been, and always will be, the future of generative AI. (SOUND ON)

396,062 views • 7 months ago

I believe general software will disappear and everyone will have custom, made for them, software. For me, I want to release more videos, but I hate editing videos. So, I spent a day vibecoding my ideal video editing software. It does exactly what I need and nothing more.

I believe general software will disappear and everyone will have custom, made for them, software. For me, I want to release more videos, but I hate editing videos. So, I spent a day vibecoding my ideal video editing software. It does exactly what I need and nothing more.

16,203 views • 16 days ago

I have playing with Black Forest Labs Black Forest Labs FLUX 3 for a few days now. (Thank you for the early access!) And I played with it... a lot. Some may say, too much. 😅 Here are some of my favorite gens. Needless to say, I love this model. Audio on! More in 🧵

I have playing with Black Forest Labs Black Forest Labs FLUX 3 for a few days now. (Thank you for the early access!) And I played with it... a lot. Some may say, too much. 😅 Here are some of my favorite gens. Needless to say, I love this model. Audio on! More in 🧵

32,042 views • 3 months ago

I trained an ACEStep 1.5 XL LoRA on "some obscure 60s English rock band". Then I wrote a song about LoRA training and had them play it. Absolutely wonderful experience. I still have some UI work before I can make training public in AI Toolkit, but working on it as fast as I can.

I trained an ACEStep 1.5 XL LoRA on "some obscure 60s English rock band". Then I wrote a song about LoRA training and had them play it. Absolutely wonderful experience. I still have some UI work before I can make training public in AI Toolkit, but working on it as fast as I can.

37,730 views • 6 months ago

Testing Flux.1 schnell embedding interpolation. I told claude some details about my life and had it generate 40 prompts from being born -> being dead using the data I gave it. Then interpolated between the embeddings for this video. Uses Araminta softpasty LoRA

Testing Flux.1 schnell embedding interpolation. I told claude some details about my life and had it generate 40 prompts from being born -> being dead using the data I gave it. Then interpolated between the embeddings for this video. Uses Araminta softpasty LoRA

45,773 views • 2 years ago

AI Toolkit now officially supports training Lightricks LTX-2 LoRAs.

AI Toolkit now officially supports training Lightricks LTX-2 LoRAs.

18,776 views • 9 months ago

I added the ability to merge LoRAs in AI Toolkit. I always find a merge of the last few checkpoints yields better results. Basically a "poor man's EMA". But more importantly, this change adds the framework to run small scripts from the frontend, which I have a lot of plans for.

I added the ability to merge LoRAs in AI Toolkit. I always find a merge of the last few checkpoints yields better results. Basically a "poor man's EMA". But more importantly, this change adds the framework to run small scripts from the frontend, which I have a lot of plans for.

11,869 views • 5 months ago

I trained another one on some obscure 90s English rock band. Same prompt and song, different performers. Someone needs to build desktop software that generates lyrics by shuffling customizable prompts, shuffles and mixes custom LoRAs, and queues up songs.

I trained another one on some obscure 90s English rock band. Same prompt and song, different performers. Someone needs to build desktop software that generates lyrics by shuffling customizable prompts, shuffles and mixes custom LoRAs, and queues up songs.

10,894 views • 6 months ago

LTX-2 by Lightricks is amazing.

LTX-2 by Lightricks is amazing.

12,109 views • 9 months ago

Wan 2.1 14B LoRA training working on 24GB! 🥳 Avg 1.7sec/step at 480p on a 4090. LR was too low on this one.

Wan 2.1 14B LoRA training working on 24GB! 🥳 Avg 1.7sec/step at 480p on a 4090. LR was too low on this one.

16,113 views • 1 year ago

No more content to load