jimmycarter

/

qwen-image-int8-quanto

like 0

• "transformers.models.qwen2_5_vl.modeling_qwen2_5_vl.Qwen2_5_VLVisionBlock",
• "torch._utils._rebuild_parameter",
• "transformers.models.qwen2_5_vl.modeling_qwen2_5_vl.Qwen2_5_VLRotaryEmbedding",
• "torch.nn.modules.conv.Conv3d",
• "torch._utils._rebuild_wrapper_subclass",
• "collections.OrderedDict",
• "torch._tensor._rebuild_from_type_v2",
• "torch.bfloat16",
• "__builtin__.set",
• "torch.nn.modules.activation.SiLU",
• "torch.serialization._get_layout",
• "torch.nn.modules.sparse.Embedding",
• "transformers.models.qwen2_5_vl.modeling_qwen2_5_vl.Qwen2_5_VLTextModel",
• "torch._utils._rebuild_tensor_v2",
• "transformers.models.qwen2_5_vl.modeling_qwen2_5_vl.Qwen2_5_VLVisionAttention",
• "optimum.quanto.tensor.qtype.qtype",
• "transformers.models.qwen2_5_vl.modeling_qwen2_5_vl.Qwen2_5_VLPatchMerger",
• "transformers.models.qwen2_5_vl.modeling_qwen2_5_vl.Qwen2RMSNorm",
• "transformers.models.qwen2_5_vl.modeling_qwen2_5_vl.Qwen2_5_VLDecoderLayer",
• "transformers.models.qwen2_5_vl.modeling_qwen2_5_vl.Qwen2_5_VLMLP",
• "optimum.quanto.tensor.weights.qbytes.WeightQBytesTensor",
• "torch.nn.modules.container.Sequential",
• "optimum.quanto.tensor.optimizers.absmax_optimizer.AbsmaxOptimizer",
• "transformers.models.qwen2_5_vl.configuration_qwen2_5_vl.Qwen2_5_VLVisionConfig",
• "transformers.models.qwen2_5_vl.modeling_qwen2_5_vl.Qwen2_5_VisionTransformerPretrainedModel",
• "transformers.models.qwen2_5_vl.configuration_qwen2_5_vl.Qwen2_5_VLTextConfig",
• "transformers.models.qwen2_5_vl.modeling_qwen2_5_vl.Qwen2_5_VLModel",
• "torch.CharStorage",
• "torch.BFloat16Storage",
• "transformers.models.qwen2_5_vl.configuration_qwen2_5_vl.Qwen2_5_VLConfig",
• "torch.nn.modules.container.ModuleList",
• "transformers.models.qwen2_5_vl.modeling_qwen2_5_vl.Qwen2_5_VLForConditionalGeneration",
• "torch.int8",
• "torch.nn.modules.activation.GELU",
• "transformers.models.qwen2_5_vl.modeling_qwen2_5_vl.Qwen2_5_VisionPatchEmbed",
• "transformers.models.qwen2_5_vl.modeling_qwen2_5_vl.Qwen2_5_VisionRotaryEmbedding",
• "transformers.modeling_rope_utils._compute_default_rope_parameters",
• "optimum.quanto.nn.qlinear.QLinear",
• "transformers.generation.configuration_utils.GenerationConfig",
• "torch.device",
• "torch.FloatStorage",
• "transformers.models.qwen2_5_vl.modeling_qwen2_5_vl.Qwen2_5_VLAttention",
• "transformers.models.qwen2_5_vl.modeling_qwen2_5_vl.Qwen2MLP"

How to fix it?