Duplicated from  camenduru/FLUX.1-schnell-quantize

UrNinja

/

FLUX.1-schnell-quantize

Like 0

• "torch._tensor._rebuild_from_type_v2",
• "torch.utils.hooks.RemovableHandle",
• "transformers.models.clip.modeling_clip.CLIPTextTransformer",
• "torch.device",
• "torch._utils._rebuild_parameter",
• "torch._utils._rebuild_tensor_v2",
• "transformers.models.clip.modeling_clip.CLIPTextModel",
• "optimum.quanto.nn.qlinear.QLinear",
• "optimum.quanto.tensor.weights.qbytes.WeightQBytesTensor",
• "transformers.models.clip.modeling_clip.CLIPEncoder",
• "torch.float8_e4m3fn",
• "transformers.models.clip.modeling_clip.CLIPEncoderLayer",
• "__builtin__.getattr",
• "transformers.models.clip.modeling_clip.CLIPMLP",
• "transformers.models.clip.modeling_clip.CLIPTextEmbeddings",
• "optimum.quanto.tensor.qtype.qtype",
• "optimum.quanto.tensor.optimizers.absmax_optimizer.AbsmaxOptimizer",
• "torch.serialization._get_layout",
• "transformers.activations.QuickGELUActivation",
• "__builtin__.set",
• "torch.BFloat16Storage",
• "collections.OrderedDict",
• "torch.nn.modules.container.ModuleList",
• "torch.nn.modules.sparse.Embedding",
• "transformers.models.clip.configuration_clip.CLIPTextConfig",
• "torch._utils._rebuild_tensor_v3",
• "transformers.models.clip.modeling_clip.CLIPSdpaAttention",
• "torch.bfloat16",
• "torch.LongStorage",
• "torch._utils._rebuild_wrapper_subclass",
• "optimum.quanto.nn.qlayernorm.QLayerNorm",
• "torch.storage.UntypedStorage"

How to fix it?

• "torch._tensor._rebuild_from_type_v2",
• "torch.utils.hooks.RemovableHandle",
• "torch.device",
• "torch._utils._rebuild_parameter",
• "torch._utils._rebuild_tensor_v2",
• "transformers.models.t5.modeling_t5.T5LayerFF",
• "optimum.quanto.nn.qlinear.QLinear",
• "optimum.quanto.tensor.weights.qbytes.WeightQBytesTensor",
• "torch.float8_e4m3fn",
• "transformers.models.t5.configuration_t5.T5Config",
• "__builtin__.getattr",
• "optimum.quanto.tensor.qtype.qtype",
• "optimum.quanto.tensor.optimizers.absmax_optimizer.AbsmaxOptimizer",
• "torch.serialization._get_layout",
• "transformers.models.t5.modeling_t5.T5Stack",
• "__builtin__.set",
• "transformers.models.t5.modeling_t5.T5Block",
• "torch.BFloat16Storage",
• "collections.OrderedDict",
• "torch.nn.modules.container.ModuleList",
• "torch.nn.modules.sparse.Embedding",
• "torch._utils._rebuild_tensor_v3",
• "torch.bfloat16",
• "torch.nn.modules.dropout.Dropout",
• "transformers.models.t5.modeling_t5.T5DenseGatedActDense",
• "transformers.models.t5.modeling_t5.T5LayerSelfAttention",
• "torch._utils._rebuild_wrapper_subclass",
• "transformers.models.t5.modeling_t5.T5EncoderModel",
• "transformers.models.t5.modeling_t5.T5LayerNorm",
• "transformers.activations.NewGELUActivation",
• "transformers.models.t5.modeling_t5.T5Attention",
• "torch.storage.UntypedStorage"

How to fix it?

• "torch._tensor._rebuild_from_type_v2",
• "torch.utils.hooks.RemovableHandle",
• "torch.device",
• "torch._utils._rebuild_parameter",
• "torch.FloatStorage",
• "diffusers.models.embeddings.Timesteps",
• "torch._utils._rebuild_tensor_v2",
• "optimum.quanto.nn.qlinear.QLinear",
• "diffusers.models.normalization.AdaLayerNormZero",
• "optimum.quanto.tensor.weights.qbytes.WeightQBytesTensor",
• "diffusers.models.attention_processor.FluxAttnProcessor2_0",
• "torch.float8_e4m3fn",
• "torch.nn.modules.activation.GELU",
• "diffusers.models.embeddings.FluxPosEmbed",
• "diffusers.models.embeddings.PixArtAlphaTextProjection",
• "diffusers.models.normalization.AdaLayerNormContinuous",
• "__builtin__.getattr",
• "optimum.quanto.tensor.qtype.qtype",
• "optimum.quanto.tensor.optimizers.absmax_optimizer.AbsmaxOptimizer",
• "diffusers.models.activations.GELU",
• "torch.serialization._get_layout",
• "diffusers.models.transformers.transformer_flux.FluxSingleTransformerBlock",
• "diffusers.models.embeddings.CombinedTimestepTextProjEmbeddings",
• "diffusers.models.embeddings.TimestepEmbedding",
• "__builtin__.set",
• "torch.BFloat16Storage",
• "diffusers.models.attention.FeedForward",
• "collections.OrderedDict",
• "diffusers.models.normalization.RMSNorm",
• "torch.nn.modules.container.ModuleList",
• "torch.Size",
• "torch._utils._rebuild_tensor_v3",
• "diffusers.models.normalization.AdaLayerNormZeroSingle",
• "diffusers.models.transformers.transformer_flux.FluxTransformerBlock",
• "torch.bfloat16",
• "torch.nn.modules.dropout.Dropout",
• "torch.nn.modules.activation.SiLU",
• "torch._utils._rebuild_wrapper_subclass",
• "optimum.quanto.nn.qlayernorm.QLayerNorm",
• "diffusers.configuration_utils.FrozenDict",
• "diffusers.models.transformers.transformer_flux.FluxTransformer2DModel",
• "diffusers.models.attention_processor.Attention",
• "torch.storage.UntypedStorage"

How to fix it?

• "torch._tensor._rebuild_from_type_v2",
• "torch.utils.hooks.RemovableHandle",
• "torch.device",
• "torch._utils._rebuild_parameter",
• "torch._utils._rebuild_tensor_v2",
• "optimum.quanto.nn.qlinear.QLinear",
• "optimum.quanto.tensor.weights.qbytes.WeightQBytesTensor",
• "torch.float8_e4m3fn",
• "torch.nn.modules.normalization.GroupNorm",
• "diffusers.models.downsampling.Downsample2D",
• "__builtin__.getattr",
• "diffusers.models.unets.unet_2d_blocks.UpDecoderBlock2D",
• "optimum.quanto.tensor.qtype.qtype",
• "optimum.quanto.tensor.optimizers.absmax_optimizer.AbsmaxOptimizer",
• "torch.serialization._get_layout",
• "diffusers.models.resnet.ResnetBlock2D",
• "diffusers.models.autoencoders.vae.Encoder",
• "diffusers.models.unets.unet_2d_blocks.DownEncoderBlock2D",
• "diffusers.models.upsampling.Upsample2D",
• "__builtin__.set",
• "torch.BFloat16Storage",
• "collections.OrderedDict",
• "torch.nn.modules.container.ModuleList",
• "torch._utils._rebuild_tensor_v3",
• "diffusers.models.autoencoders.vae.Decoder",
• "torch.bfloat16",
• "diffusers.models.autoencoders.autoencoder_kl.AutoencoderKL",
• "torch.nn.modules.dropout.Dropout",
• "torch.nn.modules.activation.SiLU",
• "optimum.quanto.nn.qconv2d.QConv2d",
• "torch._utils._rebuild_wrapper_subclass",
• "diffusers.configuration_utils.FrozenDict",
• "diffusers.models.attention_processor.Attention",
• "diffusers.models.unets.unet_2d_blocks.UNetMidBlock2D",
• "diffusers.models.attention_processor.AttnProcessor2_0",
• "torch.storage.UntypedStorage"

How to fix it?