PrunaAI

/

Orkhan-llama-2-7b-absa-QUANTO-float8bit-smashed

like 0

• "transformers.models.llama.modeling_llama.LlamaForCausalLM",
• "collections.OrderedDict",
• "torch.float16",
• "torch.nn.modules.container.ModuleDict",
• "torch._utils._rebuild_parameter",
• "transformers.models.llama.modeling_llama.LlamaSdpaAttention",
• "transformers.models.llama.configuration_llama.LlamaConfig",
• "torch.nn.modules.activation.SiLU",
• "__builtin__.set",
• "peft.tuners.lora.config.LoraConfig",
• "torch.device",
• "quanto.nn.qlinear.QLinear",
• "peft.utils.peft_types.PeftType",
• "torch.HalfStorage",
• "transformers.models.llama.modeling_llama.LlamaRMSNorm",
• "torch.nn.modules.container.ParameterDict",
• "peft.tuners.lora.layer.Linear",
• "transformers.models.llama.modeling_llama.LlamaRotaryEmbedding",
• "torch.float8_e4m3fn",
• "transformers.models.llama.modeling_llama.LlamaDecoderLayer",
• "transformers.generation.configuration_utils.GenerationConfig",
• "quanto.tensor.qtype.qtype",
• "transformers.models.llama.modeling_llama.LlamaMLP",
• "transformers.models.llama.modeling_llama.LlamaModel",
• "torch._utils._rebuild_tensor_v2",
• "torch.nn.modules.dropout.Dropout",
• "torch.nn.modules.container.ModuleList",
• "torch.nn.modules.sparse.Embedding",
• "torch.FloatStorage"

How to fix it?