MMattaparthy

/

ppo_test_final

like 0

• "transformers.trainer_utils.SchedulerType",
• "transformers.models.qwen2.modeling_qwen2.Qwen2ForSequenceClassification",
• "transformers.models.qwen2.modeling_qwen2.Qwen2RMSNorm",
• "transformers.training_args.OptimizerNames",
• "transformers.models.qwen2.modeling_qwen2.Qwen2MLP",
• "transformers.trainer_utils.SaveStrategy",
• "torch.device",
• "accelerate.utils.dataclasses.DistributedType",
• "__builtin__.getattr",
• "torch.nn.modules.linear.Linear",
• "transformers.modeling_rope_utils._compute_default_rope_parameters",
• "collections.OrderedDict",
• "trl.trainer.ppo_config.PPOConfig",
• "transformers.generation.configuration_utils.GenerationConfig",
• "transformers.trainer_pt_utils.AcceleratorConfig",
• "transformers.models.qwen2.modeling_qwen2.Qwen2Model",
• "torch.nn.modules.container.ModuleList",
• "transformers.activations.SiLUActivation",
• "torch.bfloat16",
• "torch._utils._rebuild_parameter",
• "transformers.models.qwen2.modeling_qwen2.Qwen2ForCausalLM",
• "torch.FloatStorage",
• "transformers.models.qwen2.modeling_qwen2.Qwen2Attention",
• "transformers.models.qwen2.modeling_qwen2.Qwen2RotaryEmbedding",
• "torch.BFloat16Storage",
• "transformers.trainer_utils.IntervalStrategy",
• "transformers.models.qwen2.configuration_qwen2.Qwen2Config",
• "torch.nn.modules.sparse.Embedding",
• "transformers.models.qwen2.modeling_qwen2.Qwen2DecoderLayer",
• "accelerate.state.PartialState",
• "transformers.trainer_utils.HubStrategy",
• "__builtin__.set",
• "torch._utils._rebuild_tensor_v2"

How to fix it?