optimum-rbln 0.9.3rc0__py3-none-any.whl → 0.9.4a2__py3-none-any.whl

optimum/rbln/diffusers/models/autoencoders/autoencoder_kl_temporal_decoder.py

@@ -0,0 +1,275 @@
+# Copyright 2025 Rebellions Inc. All rights reserved.
+
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at:
+
+#     http://www.apache.org/licenses/LICENSE-2.0
+
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+from typing import TYPE_CHECKING, Dict, List, Tuple, Union
+
+import rebel
+import torch  # noqa: I001
+from diffusers import AutoencoderKLTemporalDecoder
+from diffusers.models.autoencoders.vae import DecoderOutput
+from diffusers.models.modeling_outputs import AutoencoderKLOutput
+from transformers import PretrainedConfig
+
+from ....configuration_utils import RBLNCompileConfig
+from ....modeling import RBLNModel
+from ....utils.logging import get_logger
+from ...configurations import RBLNAutoencoderKLTemporalDecoderConfig
+from ...modeling_diffusers import RBLNDiffusionMixin
+from .vae import (
+    DiagonalGaussianDistribution,
+    RBLNRuntimeVAEDecoder,
+    RBLNRuntimeVAEEncoder,
+    _VAEEncoder,
+    _VAETemporalDecoder,
+)
+
+
+if TYPE_CHECKING:
+    from transformers import AutoFeatureExtractor, AutoProcessor, AutoTokenizer, PretrainedConfig, PreTrainedModel
+
+    from ...modeling_diffusers import RBLNDiffusionMixin, RBLNDiffusionMixinConfig
+
+logger = get_logger(__name__)
+
+
+class RBLNAutoencoderKLTemporalDecoder(RBLNModel):
+    auto_model_class = AutoencoderKLTemporalDecoder
+    hf_library_name = "diffusers"
+    _rbln_config_class = RBLNAutoencoderKLTemporalDecoderConfig
+
+    def __post_init__(self, **kwargs):
+        super().__post_init__(**kwargs)
+
+        if self.rbln_config.uses_encoder:
+            self.encoder = RBLNRuntimeVAEEncoder(runtime=self.model[0], main_input_name="x")
+        self.decoder = RBLNRuntimeVAEDecoder(runtime=self.model[-1], main_input_name="z")
+        self.image_size = self.rbln_config.image_size
+
+    @classmethod
+    def _wrap_model_if_needed(
+        cls, model: torch.nn.Module, rbln_config: RBLNAutoencoderKLTemporalDecoderConfig
+    ) -> torch.nn.Module:
+        decoder_model = _VAETemporalDecoder(model)
+        decoder_model.num_frames = rbln_config.decode_chunk_size
+        decoder_model.eval()
+
+        if rbln_config.uses_encoder:
+            encoder_model = _VAEEncoder(model)
+            encoder_model.eval()
+            return encoder_model, decoder_model
+        else:
+            return decoder_model
+
+    @classmethod
+    def get_compiled_model(
+        cls, model, rbln_config: RBLNAutoencoderKLTemporalDecoderConfig
+    ) -> Dict[str, rebel.RBLNCompiledModel]:
+        compiled_models = {}
+        if rbln_config.uses_encoder:
+            encoder_model, decoder_model = cls._wrap_model_if_needed(model, rbln_config)
+            enc_compiled_model = cls.compile(
+                encoder_model,
+                rbln_compile_config=rbln_config.compile_cfgs[0],
+                create_runtimes=rbln_config.create_runtimes,
+                device=rbln_config.device_map["encoder"],
+            )
+            compiled_models["encoder"] = enc_compiled_model
+        else:
+            decoder_model = cls._wrap_model_if_needed(model, rbln_config)
+        dec_compiled_model = cls.compile(
+            decoder_model,
+            rbln_compile_config=rbln_config.compile_cfgs[-1],
+            create_runtimes=rbln_config.create_runtimes,
+            device=rbln_config.device_map["decoder"],
+        )
+        compiled_models["decoder"] = dec_compiled_model
+
+        return compiled_models
+
+    @classmethod
+    def get_vae_sample_size(
+        cls,
+        pipe: "RBLNDiffusionMixin",
+        rbln_config: RBLNAutoencoderKLTemporalDecoderConfig,
+        return_vae_scale_factor: bool = False,
+    ) -> Tuple[int, int]:
+        sample_size = rbln_config.sample_size
+        if hasattr(pipe, "vae_scale_factor"):
+            vae_scale_factor = pipe.vae_scale_factor
+        else:
+            if hasattr(pipe.vae.config, "block_out_channels"):
+                vae_scale_factor = 2 ** (len(pipe.vae.config.block_out_channels) - 1)
+            else:
+                vae_scale_factor = 8  # vae image processor default value 8 (int)
+
+        if sample_size is None:
+            sample_size = pipe.unet.config.sample_size
+            if isinstance(sample_size, int):
+                sample_size = (sample_size, sample_size)
+            sample_size = (sample_size[0] * vae_scale_factor, sample_size[1] * vae_scale_factor)
+
+        if return_vae_scale_factor:
+            return sample_size, vae_scale_factor
+        else:
+            return sample_size
+
+    @classmethod
+    def update_rbln_config_using_pipe(
+        cls, pipe: "RBLNDiffusionMixin", rbln_config: "RBLNDiffusionMixinConfig", submodule_name: str
+    ) -> "RBLNDiffusionMixinConfig":
+        rbln_config.vae.sample_size, rbln_config.vae.vae_scale_factor = cls.get_vae_sample_size(
+            pipe, rbln_config.vae, return_vae_scale_factor=True
+        )
+
+        if rbln_config.vae.num_frames is None:
+            if hasattr(pipe.unet.config, "num_frames"):
+                rbln_config.vae.num_frames = pipe.unet.config.num_frames
+            else:
+                raise ValueError("num_frames should be specified in unet config.json")
+
+        if rbln_config.vae.decode_chunk_size is None:
+            rbln_config.vae.decode_chunk_size = rbln_config.vae.num_frames
+
+        def chunk_frame(num_frames, decode_chunk_size):
+            # get closest divisor to num_frames
+            divisors = [i for i in range(1, num_frames) if num_frames % i == 0]
+            closest = min(divisors, key=lambda x: abs(x - decode_chunk_size))
+            if decode_chunk_size != closest:
+                logger.warning(
+                    f"To ensure successful model compilation and prevent device OOM, {decode_chunk_size} is set to {closest}."
+                )
+            return closest
+
+        decode_chunk_size = chunk_frame(rbln_config.vae.num_frames, rbln_config.vae.decode_chunk_size)
+        rbln_config.vae.decode_chunk_size = decode_chunk_size
+        return rbln_config
+
+    @classmethod
+    def _update_rbln_config(
+        cls,
+        preprocessors: Union["AutoFeatureExtractor", "AutoProcessor", "AutoTokenizer"],
+        model: "PreTrainedModel",
+        model_config: "PretrainedConfig",
+        rbln_config: RBLNAutoencoderKLTemporalDecoderConfig,
+    ) -> RBLNAutoencoderKLTemporalDecoderConfig:
+        if rbln_config.sample_size is None:
+            rbln_config.sample_size = model_config.sample_size
+
+        if rbln_config.vae_scale_factor is None:
+            if hasattr(model_config, "block_out_channels"):
+                rbln_config.vae_scale_factor = 2 ** (len(model_config.block_out_channels) - 1)
+            else:
+                # vae image processor default value 8 (int)
+                rbln_config.vae_scale_factor = 8
+
+        compile_cfgs = []
+        if rbln_config.uses_encoder:
+            vae_enc_input_info = [
+                (
+                    "x",
+                    [
+                        rbln_config.batch_size,
+                        model_config.in_channels,
+                        rbln_config.sample_size[0],
+                        rbln_config.sample_size[1],
+                    ],
+                    "float32",
+                )
+            ]
+            compile_cfgs.append(RBLNCompileConfig(compiled_model_name="encoder", input_info=vae_enc_input_info))
+
+        decode_batch_size = rbln_config.batch_size * rbln_config.decode_chunk_size
+        vae_dec_input_info = [
+            (
+                "z",
+                [
+                    decode_batch_size,
+                    model_config.latent_channels,
+                    rbln_config.latent_sample_size[0],
+                    rbln_config.latent_sample_size[1],
+                ],
+                "float32",
+            )
+        ]
+        compile_cfgs.append(RBLNCompileConfig(compiled_model_name="decoder", input_info=vae_dec_input_info))
+
+        rbln_config.set_compile_cfgs(compile_cfgs)
+        return rbln_config
+
+    @classmethod
+    def _create_runtimes(
+        cls,
+        compiled_models: List[rebel.RBLNCompiledModel],
+        rbln_config: RBLNAutoencoderKLTemporalDecoderConfig,
+    ) -> List[rebel.Runtime]:
+        if len(compiled_models) == 1:
+            # decoder
+            expected_models = ["decoder"]
+        else:
+            expected_models = ["encoder", "decoder"]
+
+        if any(model_name not in rbln_config.device_map for model_name in expected_models):
+            cls._raise_missing_compiled_file_error(expected_models)
+
+        device_vals = [rbln_config.device_map[model_name] for model_name in expected_models]
+        return [
+            rebel.Runtime(
+                compiled_model,
+                tensor_type="pt",
+                device=device_val,
+                activate_profiler=rbln_config.activate_profiler,
+                timeout=rbln_config.timeout,
+            )
+            for compiled_model, device_val in zip(compiled_models, device_vals)
+        ]
+
+    def encode(
+        self, x: torch.FloatTensor, return_dict: bool = True
+    ) -> Union[AutoencoderKLOutput, Tuple[DiagonalGaussianDistribution]]:
+        """
+        Encode an input image into a latent representation.
+
+        Args:
+            x: The input image to encode.
+            return_dict:
+                Whether to return output as a dictionary. Defaults to True.
+
+        Returns:
+            The latent representation or AutoencoderKLOutput if return_dict=True
+        """
+        posterior = self.encoder.encode(x)
+
+        if not return_dict:
+            return (posterior,)
+
+        return AutoencoderKLOutput(latent_dist=posterior)
+
+    def decode(self, z: torch.FloatTensor, return_dict: bool = True) -> torch.FloatTensor:
+        """
+        Decode a latent representation into a video.
+
+        Args:
+            z: The latent representation to decode.
+            return_dict:
+                Whether to return output as a dictionary. Defaults to True.
+
+        Returns:
+            The decoded video or DecoderOutput if return_dict=True
+        """
+        decoded = self.decoder.decode(z)
+
+        if not return_dict:
+            return (decoded,)
+
+        return DecoderOutput(sample=decoded)

optimum/rbln/diffusers/models/autoencoders/vae.py

@@ -12,7 +12,7 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from typing import TYPE_CHECKING, List
+from typing import TYPE_CHECKING, List, Union
 
 import torch
 from diffusers.models.autoencoders.vae import DiagonalGaussianDistribution, IdentityDistribution
@@ -21,7 +21,7 @@ from ....utils.runtime_utils import RBLNPytorchRuntime
 
 
 if TYPE_CHECKING:
-    from diffusers import AutoencoderKL, AutoencoderKLCosmos, VQModel
+    from diffusers import AutoencoderKL, AutoencoderKLCosmos, AutoencoderKLTemporalDecoder, VQModel
 
 
 class RBLNRuntimeVAEEncoder(RBLNPytorchRuntime):
@@ -67,18 +67,37 @@ class _VAEDecoder(torch.nn.Module):
         return vae_out
 
 
+class _VAETemporalDecoder(torch.nn.Module):
+    def __init__(self, vae: "AutoencoderKLTemporalDecoder"):
+        super().__init__()
+        self.vae = vae
+        self.num_frames = None
+
+    def forward(self, z):
+        vae_out = self.vae.decode(z, num_frames=self.num_frames, return_dict=False)
+        return vae_out
+
+
 class _VAEEncoder(torch.nn.Module):
-    def __init__(self, vae: "AutoencoderKL"):
+    def __init__(self, vae: Union["AutoencoderKL", "AutoencoderKLTemporalDecoder"]):
         super().__init__()
         self.vae = vae
 
     def encode(self, x: torch.FloatTensor, return_dict: bool = True):
-        if self.use_tiling and (x.shape[-1] > self.tile_sample_min_size or x.shape[-2] > self.tile_sample_min_size):
-            return self.tiled_encode(x, return_dict=return_dict)
+        if hasattr(self, "use_tiling") and hasattr(self, "use_slicing"):
+            if self.use_tiling and (
+                x.shape[-1] > self.tile_sample_min_size or x.shape[-2] > self.tile_sample_min_size
+            ):
+                return self.tiled_encode(x, return_dict=return_dict)
+
+            if self.use_slicing and x.shape[0] > 1:
+                encoded_slices = [self.encoder(x_slice) for x_slice in x.split(1)]
+                h = torch.cat(encoded_slices)
+            else:
+                h = self.encoder(x)
+                if self.quant_conv is not None:
+                    h = self.quant_conv(h)
 
-        if self.use_slicing and x.shape[0] > 1:
-            encoded_slices = [self.encoder(x_slice) for x_slice in x.split(1)]
-            h = torch.cat(encoded_slices)
         else:
             h = self.encoder(x)
             if self.quant_conv is not None:

optimum/rbln/diffusers/models/controlnet.py

@@ -118,7 +118,7 @@ class RBLNControlNetModel(RBLNModel):
         )
 
     @classmethod
-    def wrap_model_if_needed(cls, model: torch.nn.Module, rbln_config: RBLNModelConfig) -> torch.nn.Module:
+    def _wrap_model_if_needed(cls, model: torch.nn.Module, rbln_config: RBLNModelConfig) -> torch.nn.Module:
         use_encoder_hidden_states = False
         for down_block in model.down_blocks:
             if use_encoder_hidden_states := getattr(down_block, "has_cross_attention", False):
@@ -215,10 +215,25 @@ class RBLNControlNetModel(RBLNModel):
         encoder_hidden_states: torch.Tensor,
         controlnet_cond: torch.FloatTensor,
         conditioning_scale: torch.Tensor = 1.0,
-        added_cond_kwargs: Dict[str, torch.Tensor] = {},
+        added_cond_kwargs: Optional[Dict[str, torch.Tensor]] = None,
         return_dict: bool = True,
         **kwargs,
     ):
+        """
+        Forward pass for the RBLN-optimized ControlNetModel.
+
+        Args:
+            sample (torch.FloatTensor): The noisy input tensor.
+            timestep (Union[torch.Tensor, float, int]): The number of timesteps to denoise an input.
+            encoder_hidden_states (torch.Tensor): The encoder hidden states.
+            controlnet_cond (torch.FloatTensor): The conditional input tensor of shape `(batch_size, max_seq_len, hidden_size)`.
+            conditioning_scale (torch.Tensor): The scale factor for ControlNet outputs.
+            added_cond_kwargs (Dict[str, torch.Tensor]): Additional conditions for the Stable Diffusion XL UNet.
+            return_dict (bool): Whether or not to return a [`~diffusers.models.controlnets.controlnet.ControlNetOutput`] instead of a plain tuple
+
+        Returns:
+            (Union[`~diffusers.models.controlnets.controlnet.ControlNetOutput`], Tuple)
+        """
         sample_batch_size = sample.size()[0]
         compiled_batch_size = self.compiled_batch_size
         if sample_batch_size != compiled_batch_size and (

optimum/rbln/diffusers/models/transformers/prior_transformer.py

@@ -77,7 +77,7 @@ class RBLNPriorTransformer(RBLNModel):
         self.clip_std = artifacts["clip_std"]
 
     @classmethod
-    def wrap_model_if_needed(cls, model: torch.nn.Module, rbln_config: RBLNModelConfig) -> torch.nn.Module:
+    def _wrap_model_if_needed(cls, model: torch.nn.Module, rbln_config: RBLNModelConfig) -> torch.nn.Module:
         return _PriorTransformer(model).eval()
 
     @classmethod
@@ -128,13 +128,27 @@ class RBLNPriorTransformer(RBLNModel):
 
     def forward(
         self,
-        hidden_states,
+        hidden_states: torch.Tensor,
         timestep: Union[torch.Tensor, float, int],
         proj_embedding: torch.Tensor,
         encoder_hidden_states: Optional[torch.Tensor] = None,
         attention_mask: Optional[torch.Tensor] = None,
         return_dict: bool = True,
     ):
+        """
+        Forward pass for the RBLN-optimized PriorTransformer.
+
+        Args:
+            hidden_states (torch.Tensor): The currently predicted image embeddings.
+            timestep (Union[torch.Tensor, float, int]): Current denoising step.
+            proj_embedding (torch.Tensor): Projected embedding vector the denoising process is conditioned on.
+            encoder_hidden_states (Optional[torch.Tensor]): Hidden states of the text embeddings the denoising process is conditioned on.
+            attention_mask (Optional[torch.Tensor]): Text mask for the text embeddings.
+            return_dict (bool): Whether or not to return a [`~diffusers.models.transformers.prior_transformer.PriorTransformerOutput`] instead of a plain tuple.
+
+        Returns:
+            (Union[`~diffusers.models.transformers.prior_transformer.PriorTransformerOutput`, Tuple])
+        """
         # Convert timestep(long) and attention_mask(bool) to float
         return super().forward(
             hidden_states,

optimum/rbln/diffusers/models/transformers/transformer_cosmos.py

@@ -185,7 +185,7 @@ class RBLNCosmosTransformer3DModel(RBLNModel):
         )
 
     @classmethod
-    def wrap_model_if_needed(cls, model: torch.nn.Module, rbln_config: RBLNModelConfig) -> torch.nn.Module:
+    def _wrap_model_if_needed(cls, model: torch.nn.Module, rbln_config: RBLNModelConfig) -> torch.nn.Module:
         num_latent_frames = rbln_config.num_latent_frames
         latent_height = rbln_config.latent_height
         latent_width = rbln_config.latent_width
@@ -303,6 +303,21 @@ class RBLNCosmosTransformer3DModel(RBLNModel):
         padding_mask: Optional[torch.Tensor] = None,
         return_dict: bool = True,
     ):
+        """
+        Forward pass for the RBLN-optimized CosmosTransformer3DModel.
+
+        Args:
+            hidden_states (torch.Tensor): The currently predicted image embeddings.
+            timestep (torch.Tensor): Current denoising step.
+            encoder_hidden_states (torch.Tensor): Conditional embeddings (embeddings computed from the input conditions such as prompts) to use.
+            fps: (Optional[int]): Frames per second for the video being generated.
+            condition_mask (Optional[torch.Tensor]): Tensor of condition mask.
+            padding_mask (Optional[torch.Tensor]): Tensor of padding mask.
+            return_dict (bool): Whether or not to return a [`~diffusers.models.modeling_output.Transformer2DModelOutput`] instead of a plain tuple.
+
+        Returns:
+            (Union[`~diffusers.models.modeling_output.Transformer2DModelOutput`, Tuple])
+        """
         (
             hidden_states,
             temb,

optimum/rbln/diffusers/models/transformers/transformer_sd3.py

@@ -77,7 +77,7 @@ class RBLNSD3Transformer2DModel(RBLNModel):
         super().__post_init__(**kwargs)
 
     @classmethod
-    def wrap_model_if_needed(cls, model: torch.nn.Module, rbln_config: RBLNModelConfig) -> torch.nn.Module:
+    def _wrap_model_if_needed(cls, model: torch.nn.Module, rbln_config: RBLNModelConfig) -> torch.nn.Module:
         return SD3Transformer2DModelWrapper(model).eval()
 
     @classmethod
@@ -161,6 +161,19 @@ class RBLNSD3Transformer2DModel(RBLNModel):
         return_dict: bool = True,
         **kwargs,
     ):
+        """
+        Forward pass for the RBLN-optimized SD3Transformer2DModel.
+
+        Args:
+            hidden_states (torch.FloatTensor): The currently predicted image embeddings.
+            encoder_hidden_states (torch.FloatTensor): Conditional embeddings (embeddings computed from the input conditions such as prompts) to use.
+            pooled_projections (torch.FloatTensor): Embeddings projected from the embeddings of input conditions.
+            timestep (torch.LongTensor): Current denoising step.
+            return_dict (bool): Whether or not to return a [`~diffusers.models.modeling_output.Transformer2DModelOutput`] instead of a plain tuple.
+
+        Returns:
+            (Union[`~diffusers.models.modeling_output.Transformer2DModelOutput`, Tuple])
+        """
         sample_batch_size = hidden_states.size()[0]
         compiled_batch_size = self.compiled_batch_size
         if sample_batch_size != compiled_batch_size and (

optimum/rbln/diffusers/models/unets/__init__.py

@@ -13,3 +13,4 @@
 # limitations under the License.
 
 from .unet_2d_condition import RBLNUNet2DConditionModel
+from .unet_spatio_temporal_condition import RBLNUNetSpatioTemporalConditionModel

optimum/rbln/diffusers/models/unets/unet_2d_condition.py

@@ -171,7 +171,7 @@ class RBLNUNet2DConditionModel(RBLNModel):
             self.add_embedding = ADDEMBEDDING(LINEAR1(self.in_features))
 
     @classmethod
-    def wrap_model_if_needed(
+    def _wrap_model_if_needed(
         cls, model: torch.nn.Module, rbln_config: RBLNUNet2DConditionModelConfig
     ) -> torch.nn.Module:
         if model.config.addition_embed_type == "text_time":
@@ -341,7 +341,7 @@ class RBLNUNet2DConditionModel(RBLNModel):
         timestep_cond: Optional[torch.Tensor] = None,
         attention_mask: Optional[torch.Tensor] = None,
         cross_attention_kwargs: Optional[Dict[str, Any]] = None,
-        added_cond_kwargs: Dict[str, torch.Tensor] = {},
+        added_cond_kwargs: Optional[Dict[str, torch.Tensor]] = None,
         down_block_additional_residuals: Optional[Tuple[torch.Tensor]] = None,
         mid_block_additional_residual: Optional[torch.Tensor] = None,
         down_intrablock_additional_residuals: Optional[Tuple[torch.Tensor]] = None,
@@ -349,6 +349,22 @@ class RBLNUNet2DConditionModel(RBLNModel):
         return_dict: bool = True,
         **kwargs,
     ) -> Union[UNet2DConditionOutput, Tuple]:
+        """
+        Forward pass for the RBLN-optimized UNet2DConditionModel.
+
+        Args:
+            sample (torch.Tensor): The noisy input tensor with the following shape `(batch, channel, height, width)`.
+            timestep (Union[torch.Tensor, float, int]): The number of timesteps to denoise an input.
+            encoder_hidden_states (torch.Tensor): The encoder hidden states.
+            added_cond_kwargs (Dict[str, torch.Tensor]): A kwargs dictionary containing additional embeddings that
+                if specified are added to the embeddings that are passed along to the UNet blocks.
+            down_block_additional_residuals (Optional[Tuple[torch.Tensor]]): A tuple of tensors that if specified are added to the residuals of down unet blocks.
+            mid_block_additional_residual (Optional[torch.Tensor]): A tensor that if specified is added to the residual of the middle unet block.
+            return_dict (bool): Whether or not to return a [`~diffusers.models.unets.unet_2d_condition.UNet2DConditionOutput`] instead of a plain tuple.
+
+        Returns:
+            (Union[`~diffusers.models.unets.unet_2d_condition.UNet2DConditionOutput`], Tuple)
+        """
         sample_batch_size = sample.size()[0]
         compiled_batch_size = self.compiled_batch_size
         if sample_batch_size != compiled_batch_size and (