Echoes123-3

/

moe-router-test

like 0

• "torch.bfloat16",
• "torch.FloatStorage",
• "torch.nn.modules.linear.Linear",
• "transformers.models.bert.modeling_bert.BertSdpaSelfAttention",
• "transformers.models.bert.modeling_bert.BertModel",
• "transformers.models.bert.modeling_bert.BertEncoder",
• "torch.float32",
• "transformers.quantizers.quantizer_bnb_4bit.Bnb4BitHfQuantizer",
• "transformers.models.phi.modeling_phi.PhiForCausalLM",
• "transformers.models.llama.modeling_llama.LlamaForCausalLM",
• "collections.OrderedDict",
• "torch.uint8",
• "transformers.models.bert.modeling_bert.BertPooler",
• "transformers.models.bert.configuration_bert.BertConfig",
• "torch.nn.modules.dropout.Dropout",
• "transformers.models.bert.modeling_bert.BertSelfOutput",
• "transformers.models.llama.modeling_llama.LlamaRMSNorm",
• "__builtin__.set",
• "transformers.models.bert.modeling_bert.BertIntermediate",
• "transformers.modeling_rope_utils._compute_default_rope_parameters",
• "transformers.models.bert.modeling_bert.BertAttention",
• "transformers.models.phi.modeling_phi.PhiMLP",
• "torch.Size",
• "transformers.models.phi.modeling_phi.PhiRotaryEmbedding",
• "transformers.models.llama.configuration_llama.LlamaConfig",
• "torch.LongStorage",
• "torch._C._nn.gelu",
• "transformers.models.llama.modeling_llama.LlamaModel",
• "torch.nn.modules.sparse.Embedding",
• "transformers.models.llama.modeling_llama.LlamaAttention",
• "bitsandbytes.nn.modules.Linear4bit",
• "transformers.models.phi.modeling_phi.PhiDecoderLayer",
• "transformers.models.bert.modeling_bert.BertOutput",
• "torch.ByteStorage",
• "transformers.models.llama.modeling_llama.LlamaMLP",
• "torch.nn.modules.container.ModuleList",
• "torch._utils._rebuild_tensor_v2",
• "torch._utils._rebuild_parameter",
• "transformers.models.phi.modeling_phi.PhiModel",
• "torch._utils._rebuild_parameter_with_state",
• "torch.float16",
• "transformers.utils.quantization_config.BitsAndBytesConfig",
• "transformers.models.bert.modeling_bert.BertEmbeddings",
• "transformers.models.llama.modeling_llama.LlamaDecoderLayer",
• "torch.HalfStorage",
• "torch.nn.modules.activation.Tanh",
• "transformers.models.bert.modeling_bert.BertLayer",
• "__main__.SimpleMoE",
• "transformers.activations.NewGELUActivation",
• "torch.nn.modules.normalization.LayerNorm",
• "transformers.models.llama.modeling_llama.LlamaRotaryEmbedding",
• "transformers.models.phi.configuration_phi.PhiConfig",
• "transformers.activations.GELUActivation",
• "transformers.utils.quantization_config.QuantizationMethod",
• "bitsandbytes.functional.QuantState",
• "transformers.models.phi.modeling_phi.PhiAttention",
• "transformers.generation.configuration_utils.GenerationConfig",
• "transformers.activations.SiLUActivation"

How to fix it?