diffusers/pr_14409/en/_app/immutable/nodes/45.DVuKgrIN.js

import"../chunks/DsnmJJEf.js";import{i as D,h as x,C as M,H as l,D as m,E as B,s as w}from"../chunks/BtE7mKSK.js";import{p as y,o as k,s as e,f as j,a as f,b as N,c as p,d as _,n as h,r as u}from"../chunks/jDjavuwI.js";const z='{"title":"BriaTransformer2DModel","local":"briatransformer2dmodel","sections":[{"title":"BriaTransformer2DModel","local":"diffusers.BriaTransformer2DModel","sections":[],"depth":2}],"depth":1}';var A=_('<meta name="hf:doc:metadata"/>'),F=_(`<p></p> <!> <!> <p>A modified flux Transformer model from <a href="https://huggingface.co/briaai/BRIA-3.2" rel="nofollow">Bria</a></p> <!> <div class="docstring border-l-2 border-t-2 pl-4 pt-3.5 border-gray-100 rounded-tl-xl mb-6 mt-8"><!> <p>The Transformer model introduced in Flux. Based on FluxPipeline with several changes:</p> <ul><li>no pooled embeddings</li> <li>We use zero padding for prompts</li> <li>No guidance embedding since this is not a distilled version
Reference: <a href="https://blackforestlabs.ai/announcing-black-forest-labs/" rel="nofollow">https://blackforestlabs.ai/announcing-black-forest-labs/</a></li></ul> <div class="docstring border-l-2 border-t-2 pl-4 pt-3.5 border-gray-100 rounded-tl-xl mb-6 mt-8"><!> <p>The <a href="/docs/diffusers/pr_14409/en/api/models/bria_transformer#diffusers.BriaTransformer2DModel">BriaTransformer2DModel</a> forward method.</p></div></div> <!> <p></p>`,1);function C(g,b){y(b,!1),k(()=>{new URLSearchParams(window.location.search).get("fw")}),D();var r=F();x("l3rflh",d=>{var c=A();w(c,"content",z),f(d,c)});var n=e(j(r),2);M(n,{containerStyle:"float: right; margin-left: 10px; display: inline-flex; position: relative; z-index: 10;"});var s=e(n,2);l(s,{title:"BriaTransformer2DModel",local:"briatransformer2dmodel",headingTag:"h1"});var t=e(s,4);l(t,{title:"BriaTransformer2DModel",local:"diffusers.BriaTransformer2DModel",headingTag:"h2"});var o=e(t,2),a=p(o);m(a,{name:"class diffusers.BriaTransformer2DModel",anchor:"diffusers.BriaTransformer2DModel",source:"https://github.com/huggingface/diffusers/blob/vr_14409/src/diffusers/models/transformers/transformer_bria.py#L504",parameters:[{name:"patch_size",val:": int = 1"},{name:"in_channels",val:": int = 64"},{name:"num_layers",val:": int = 19"},{name:"num_single_layers",val:": int = 38"},{name:"attention_head_dim",val:": int = 128"},{name:"num_attention_heads",val:": int = 24"},{name:"joint_attention_dim",val:": int = 4096"},{name:"pooled_projection_dim",val:": int = None"},{name:"guidance_embeds",val:": bool = False"},{name:"axes_dims_rope",val:": list = [16, 56, 56]"},{name:"rope_theta",val:" = 10000"},{name:"time_theta",val:" = 10000"}],parametersDescription:[{anchor:"diffusers.BriaTransformer2DModel.patch_size",description:"<strong>patch_size</strong> (<code>int</code>) &#x2014; Patch size to turn the input data into small patches.",name:"patch_size"},{anchor:"diffusers.BriaTransformer2DModel.in_channels",description:"<strong>in_channels</strong> (<code>int</code>, <em>optional</em>, defaults to 16) &#x2014; The number of channels in the input.",name:"in_channels"},{anchor:"diffusers.BriaTransformer2DModel.num_layers",description:"<strong>num_layers</strong> (<code>int</code>, <em>optional</em>, defaults to 18) &#x2014; The number of layers of MMDiT blocks to use.",name:"num_layers"},{anchor:"diffusers.BriaTransformer2DModel.num_single_layers",description:"<strong>num_single_layers</strong> (<code>int</code>, <em>optional</em>, defaults to 18) &#x2014; The number of layers of single DiT blocks to use.",name:"num_single_layers"},{anchor:"diffusers.BriaTransformer2DModel.attention_head_dim",description:"<strong>attention_head_dim</strong> (<code>int</code>, <em>optional</em>, defaults to 64) &#x2014; The number of channels in each head.",name:"attention_head_dim"},{anchor:"diffusers.BriaTransformer2DModel.num_attention_heads",description:"<strong>num_attention_heads</strong> (<code>int</code>, <em>optional</em>, defaults to 18) &#x2014; The number of heads to use for multi-head attention.",name:"num_attention_heads"},{anchor:"diffusers.BriaTransformer2DModel.joint_attention_dim",description:"<strong>joint_attention_dim</strong> (<code>int</code>, <em>optional</em>) &#x2014; The number of <code>encoder_hidden_states</code> dimensions to use.",name:"joint_attention_dim"},{anchor:"diffusers.BriaTransformer2DModel.pooled_projection_dim",description:"<strong>pooled_projection_dim</strong> (<code>int</code>) &#x2014; Number of dimensions to use when projecting the <code>pooled_projections</code>.",name:"pooled_projection_dim"},{anchor:"diffusers.BriaTransformer2DModel.guidance_embeds",description:"<strong>guidance_embeds</strong> (<code>bool</code>, defaults to False) &#x2014; Whether to use guidance embeddings.",name:"guidance_embeds"}]});var i=e(a,6),T=p(i);m(T,{name:"forward",anchor:"diffusers.BriaTransformer2DModel.forward",source:"https://github.com/huggingface/diffusers/blob/vr_14409/src/diffusers/models/transformers/transformer_bria.py#L582",parameters:[{name:"hidden_states",val:": Tensor"},{name:"encoder_hidden_states",val:": Tensor = None"},{name:"pooled_projections",val:": Tensor = None"},{name:"timestep",val:": LongTensor = None"},{name:"img_ids",val:": Tensor = None"},{name:"txt_ids",val:": Tensor = None"},{name:"guidance",val:": Tensor = None"},{name:"attention_kwargs",val:": dict[str, typing.Any] | None = None"},{name:"return_dict",val:": bool = True"},{name:"controlnet_block_samples",val:" = None"},{name:"controlnet_single_block_samples",val:" = None"}],parametersDescription:[{anchor:"diffusers.BriaTransformer2DModel.forward.hidden_states",description:`<strong>hidden_states</strong> (<code>torch.FloatTensor</code> of shape <code>(batch size, channel, height, width)</code>) &#x2014;
Input <code>hidden_states</code>.`,name:"hidden_states"},{anchor:"diffusers.BriaTransformer2DModel.forward.encoder_hidden_states",description:`<strong>encoder_hidden_states</strong> (<code>torch.FloatTensor</code> of shape <code>(batch size, sequence_len, embed_dims)</code>) &#x2014;
Conditional embeddings (embeddings computed from the input conditions such as prompts) to use.`,name:"encoder_hidden_states"},{anchor:"diffusers.BriaTransformer2DModel.forward.pooled_projections",description:`<strong>pooled_projections</strong> (<code>torch.FloatTensor</code> of shape <code>(batch_size, projection_dim)</code>) &#x2014; Embeddings projected
from the embeddings of input conditions.`,name:"pooled_projections"},{anchor:"diffusers.BriaTransformer2DModel.forward.timestep",description:`<strong>timestep</strong> ( <code>torch.LongTensor</code>) &#x2014;
Used to indicate denoising step.`,name:"timestep"},{anchor:"diffusers.BriaTransformer2DModel.forward.img_ids",description:`<strong>img_ids</strong> (<code>torch.Tensor</code>) &#x2014;
Image position ids used to compute the rotary positional embeddings.`,name:"img_ids"},{anchor:"diffusers.BriaTransformer2DModel.forward.txt_ids",description:`<strong>txt_ids</strong> (<code>torch.Tensor</code>) &#x2014;
Text position ids used to compute the rotary positional embeddings.`,name:"txt_ids"},{anchor:"diffusers.BriaTransformer2DModel.forward.guidance",description:`<strong>guidance</strong> (<code>torch.Tensor</code>, <em>optional</em>) &#x2014;
Guidance scale embedding used for guidance-distilled variants of the model.`,name:"guidance"},{anchor:"diffusers.BriaTransformer2DModel.forward.controlnet_block_samples",description:`<strong>controlnet_block_samples</strong> (<code>list</code> of <code>torch.Tensor</code>, <em>optional</em>) &#x2014;
A list of tensors that if specified are added to the residuals of transformer blocks.`,name:"controlnet_block_samples"},{anchor:"diffusers.BriaTransformer2DModel.forward.controlnet_single_block_samples",description:`<strong>controlnet_single_block_samples</strong> (<code>list</code> of <code>torch.Tensor</code>, <em>optional</em>) &#x2014;
A list of tensors that if specified are added to the residuals of single transformer blocks.`,name:"controlnet_single_block_samples"},{anchor:"diffusers.BriaTransformer2DModel.forward.attention_kwargs",description:`<strong>attention_kwargs</strong> (<code>dict</code>, <em>optional</em>) &#x2014;
A kwargs dictionary that if specified is passed along to the <code>AttentionProcessor</code> as defined under
<code>self.processor</code> in
<a href="https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/attention_processor.py" rel="nofollow">diffusers.models.attention_processor</a>.`,name:"attention_kwargs"},{anchor:"diffusers.BriaTransformer2DModel.forward.return_dict",description:`<strong>return_dict</strong> (<code>bool</code>, <em>optional</em>, defaults to <code>True</code>) &#x2014;
Whether or not to return a <code>~models.transformer_2d.Transformer2DModelOutput</code> instead of a plain
tuple.`,name:"return_dict"}],returnDescription:`<script context="module">export const metadata = 'undefined';<\/script>


<p>If <code>return_dict</code> is True, an <code>~models.transformer_2d.Transformer2DModelOutput</code> is returned, otherwise a
<code>tuple</code> where the first element is the sample tensor.</p>
`}),h(2),u(i),u(o);var v=e(o,2);B(v,{source:"https://github.com/huggingface/diffusers/blob/main/docs/source/en/api/models/bria_transformer.md"}),h(2),f(g,r),N()}export{C as component};