Cosmos predict2 model example.

Author: comfyanonymous

README.md

@@ -70,6 +70,8 @@ Here are some more advanced examples:
 
 [Nvidia Cosmos](cosmos)
 
+[Nvidia Cosmos Predict2](cosmos_predict2)
+
 [Wan](wan)
 
 [Audio Models](audio)

cosmos/README.md

@@ -1,4 +1,6 @@
-# Nvidia Cosmos Models
+# Original Nvidia Cosmos Models
+
+For the newer Cosmos models see [Cosmos Predict2](../cosmos_predict2)
 
 [Nvidia Cosmos](https://www.nvidia.com/en-us/ai/cosmos/) is a family of "World Models". ComfyUI currently supports specifically the 7B and 14B text to video diffusion models and the 7B and 14B image to video diffusion models.
 

cosmos_predict2/README.md

@@ -0,0 +1,46 @@
+# Nvidia Cosmos Predict2
+
+These are a family of text to image and image to video models from Nvidia.
+
+## Files to Download
+
+You will first need:
+
+#### Text encoder and VAE:
+
+[oldt5_xxl_fp8_e4m3fn_scaled.safetensors](https://huggingface.co/comfyanonymous/cosmos_1.0_text_encoder_and_VAE_ComfyUI/tree/main/text_encoders) goes in: ComfyUI/models/text_encoders/
+
+[wan_2.1_vae.safetensors](https://huggingface.co/Comfy-Org/Wan_2.1_ComfyUI_repackaged/blob/main/split_files/vae/wan_2.1_vae.safetensors) goes in: ComfyUI/models/vae/
+
+
+Note: oldt5_xxl is not the same as the t5xxl used in flux and other models. 
+oldt5_xxl is t5xxl 1.0 while the one used in flux and others is t5xxl 1.1
+
+
+You can find all the diffusion models (go in ComfyUI/models/diffusion_models/) here: [Repackaged safetensors files](https://huggingface.co/Comfy-Org/Cosmos_Predict2_repackaged/tree/main) or [Official Nvidia Model Files](https://huggingface.co/collections/nvidia/cosmos-predict2-68028efc052239369a0f2959)
+
+
+## Workflows
+
+### Text to Image
+
+This workflow uses the 2B text to image cosmos predict2 model. The file used in the workflow is [cosmos_predict2_2B_t2i.safetensors](https://huggingface.co/Comfy-Org/Cosmos_Predict2_repackaged/blob/main/cosmos_predict2_2B_t2i.safetensors) this file goes in: ComfyUI/models/diffusion_models/
+
+![Example](cosmos_predict2_2b_t2i_example.png)
+
+You can load this image in [ComfyUI](https://github.com/comfyanonymous/ComfyUI) to get the full workflow.
+
+I think the 2B model is the most interesting one but you can find the bigger 14B model here: [cosmos_predict2_14B_t2i.safetensors](https://huggingface.co/Comfy-Org/Cosmos_Predict2_repackaged/blob/main/cosmos_predict2_14B_t2i.safetensors) and use it in the workflow above.
+
+
+### Image to Video
+
+These models are pretty picky about the resolution/length of the videos. This workflow is for the 480p models, for the 720p models you will have to set the resolution to 720p or your results might be bad.
+
+This workflow uses the 2B image to video cosmos predict2 model. The file used in the workflow is [cosmos_predict2_2B_video2world_480p_16fps.safetensors](https://huggingface.co/Comfy-Org/Cosmos_Predict2_repackaged/blob/main/cosmos_predict2_2B_video2world_480p_16fps.safetensors) this file goes in: ComfyUI/models/diffusion_models/
+
+![Example](cosmos_predict2_2b_i2v_example.webp)
+
+[Workflow in Json format](cosmos_predict2_2b_i2v_example.json)
+
+