CogView4 Control Block #10809
Open
CogView4 Control Block #10809
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…diffusers into cogview4_control
yiyixuxu
reviewed
Feb 20, 2025
| @@ -35,7 +36,7 @@ class CogView4PatchEmbed(nn.Module): | |||
| def __init__( | |||
| self, | |||
| in_channels: int = 16, | |||
| hidden_size: int = 2560, | |||
| hidden_size: int = 4096, | |||
There was a problem hiding this comment.
is this not breaking change?
| """ | ||
|
|
||
|
|
||
| def calculate_shift( |
There was a problem hiding this comment.
we can add a #Copied from here
| >>> import torch | ||
| >>> from diffusers import CogView4Pipeline | ||
|
|
||
| >>> pipe = CogView4Pipeline.from_pretrained("THUDM/CogView4-6B", torch_dtype=torch.bfloat16) |
There was a problem hiding this comment.
need to update the pipeline
do we have a checkpoint?
| self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1) if getattr(self, "vae", None) else 8 | ||
| self.image_processor = VaeImageProcessor(vae_scale_factor=self.vae_scale_factor) | ||
|
|
||
| def _get_glm_embeds( |
| prompt_embeds = prompt_embeds.view(batch_size * num_images_per_prompt, seq_len, -1) | ||
| return prompt_embeds | ||
|
|
||
| def encode_prompt( |
| if timesteps is None | ||
| else np.array(timesteps) | ||
| ) | ||
| timesteps = timesteps.astype(np.int64) |
There was a problem hiding this comment.
Suggested change
| timesteps = timesteps.astype(np.int64) | |
| timesteps = timesteps.astype(np.int64).astype(np.float32) |
| self.scheduler.config.get("base_shift", 0.25), | ||
| self.scheduler.config.get("max_shift", 0.75), | ||
| ) | ||
| _, num_inference_steps = retrieve_timesteps(self.scheduler, num_inference_steps, device, sigmas=sigmas, mu=mu) |
There was a problem hiding this comment.
Suggested change
| _, num_inference_steps = retrieve_timesteps(self.scheduler, num_inference_steps, device, sigmas=sigmas, mu=mu) | |
| timesteps, num_inference_steps = retrieve_timesteps(self.scheduler, num_inference_steps, device, timesteps, sigmas=sigmas, mu=mu) |
we updated our scheduler to work with cogview4 - is there any reason we still cannot use the 1scheduler.set_timesteps1 to set timesteps?
| self.scheduler.config.get("max_shift", 0.75), | ||
| ) | ||
| _, num_inference_steps = retrieve_timesteps(self.scheduler, num_inference_steps, device, sigmas=sigmas, mu=mu) | ||
| timesteps = torch.from_numpy(timesteps).to(device) |
There was a problem hiding this comment.
Suggested change
| timesteps = torch.from_numpy(timesteps).to(device) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
What does this pull request do?
The purpose of this PR is to add a Control module to CogView4, which refers to the implementation of Flux.
Who can review?
@arrow