PrunaAI

/

microsoft-phi-2-QUANTO-float8bit-smashed

like 0

Follow

Pruna AI 243

• "torch._utils._rebuild_tensor_v2",
• "transformers.models.phi.modeling_phi.PhiModel",
• "torch._utils._rebuild_parameter",
• "torch.nn.modules.normalization.LayerNorm",
• "torch.nn.modules.dropout.Dropout",
• "transformers.models.phi.modeling_phi.PhiSdpaAttention",
• "torch.device",
• "torch.float16",
• "quanto.tensor.qtype.qtype",
• "torch.nn.modules.sparse.Embedding",
• "torch.FloatStorage",
• "quanto.nn.qlinear.QLinear",
• "transformers.activations.NewGELUActivation",
• "transformers.models.phi.modeling_phi.PhiMLP",
• "transformers.generation.configuration_utils.GenerationConfig",
• "torch.nn.modules.container.ModuleList",
• "torch.HalfStorage",
• "__builtin__.set",
• "transformers.models.phi.modeling_phi.PhiRotaryEmbedding",
• "torch.float8_e4m3fn",
• "collections.OrderedDict",
• "transformers.models.phi.configuration_phi.PhiConfig",
• "transformers.models.phi.modeling_phi.PhiForCausalLM",
• "transformers.models.phi.modeling_phi.PhiDecoderLayer"

How to fix it?