diff --git a/colossalai/shardformer/policies/auto_policy.py b/colossalai/shardformer/policies/auto_policy.py
index 640b61b579bd..0ad9a3e95a0e 100644
--- a/colossalai/shardformer/policies/auto_policy.py
+++ b/colossalai/shardformer/policies/auto_policy.py
@@ -45,7 +45,7 @@ class PolicyLocation:
 
     # LLaMA
     "transformers.models.llama.modeling_llama.LlamaModel":
-        PolicyLocation(file_name="llama", class_name="LlamaPolicy"),
+        PolicyLocation(file_name="llama", class_name="LlamaModelPolicy"),
     "transformers.models.llama.modeling_llama.LlamaForCausalLM":
         PolicyLocation(file_name="llama", class_name="LlamaForCausalLMPolicy"),
     "transformers.models.llama.modeling_llama.LlamaForSequenceClassification":
diff --git a/colossalai/shardformer/policies/bert.py b/colossalai/shardformer/policies/bert.py
index eacd0b449ad4..2b2c003ffb04 100644
--- a/colossalai/shardformer/policies/bert.py
+++ b/colossalai/shardformer/policies/bert.py
@@ -193,7 +193,7 @@ def get_held_layers(self) -> List[Module]:
         module = self.model
         stage_manager = self.pipeline_stage_manager
         held_layers = []
-        layers_per_stage = self.distribute_layers(len(self.model.encoder.layer), stage_manager.num_stages)
+        layers_per_stage = self.distribute_layers(len(module.encoder.layer), stage_manager.num_stages)
         if stage_manager.is_first_stage():
             held_layers.append(module.embeddings)
         start_idx, end_idx = self.get_stage_index(layers_per_stage, stage_manager.stage)
diff --git a/colossalai/shardformer/policies/llama.py b/colossalai/shardformer/policies/llama.py
index b10e07560d22..b2b6470188a4 100644
--- a/colossalai/shardformer/policies/llama.py
+++ b/colossalai/shardformer/policies/llama.py
@@ -1,11 +1,30 @@
-from typing import Dict, Union
+import math
+from functools import partial
+from types import MethodType
+from typing import Dict, List, Optional, Tuple, Union
 
+import torch
 import torch.nn as nn
+from torch import Tensor
+from torch.nn import BCEWithLogitsLoss, CrossEntropyLoss, Module, MSELoss
+from transformers.modeling_outputs import (
+    BaseModelOutputWithPast,
+    BaseModelOutputWithPastAndCrossAttentions,
+    BaseModelOutputWithPoolingAndCrossAttentions,
+    CausalLMOutputWithCrossAttentions,
+    CausalLMOutputWithPast,
+    SequenceClassifierOutputWithPast,
+)
+from transformers.models.llama.modeling_llama import LlamaForCausalLM, LlamaForSequenceClassification, LlamaModel
+from transformers.utils import ModelOutput, logging
 
+from colossalai.pipeline.stage_manager import PipelineStageManager
 from colossalai.shardformer.layer import FusedRMSNorm, Linear1D_Col, Linear1D_Row, VocabParallelEmbedding1D
 
 from .base_policy import ModulePolicyDescription, Policy, SubModuleReplacementDescription
 
+logger = logging.get_logger(__name__)
+
 __all__ = ['LlamaPolicy', 'LlamaForCausalLMPolicy', 'LlamaForSequenceClassificationPolicy']
 
 
@@ -106,6 +125,43 @@ def postprocess(self):
         return self.model
 
 
+class LlamaModelPolicy(LlamaPolicy):
+
+    def __init__(self) -> None:
+        super().__init__()
+
+    def module_policy(self):
+        module_policy = super().module_policy()
+        from transformers.models.llama.modeling_llama import LlamaModel
+        if self.pipeline_stage_manager:
+            # set None as default
+            stage_manager = self.pipeline_stage_manager
+            layers_per_stage = Policy.distribute_layers(len(self.model.layers), stage_manager.num_stages)
+            stage_index = Policy.get_stage_index(layers_per_stage, stage_manager.stage)
+            module_policy[LlamaModel] = ModulePolicyDescription(method_replacement={
+                'forward': partial(llama_model_forward, stage_manager=stage_manager, stage_index=stage_index)
+            })
+        return module_policy
+
+    def get_held_layers(self) -> List[Module]:
+        """Get pipeline layers for current stage."""
+        module = self.model
+        stage_manager = self.pipeline_stage_manager
+        held_layers = []
+        layers_per_stage = self.distribute_layers(len(module.layers), stage_manager.num_stages)
+        if stage_manager.is_first_stage():
+            held_layers.append(module.embed_tokens)
+        start_idx, end_idx = self.get_stage_index(layers_per_stage, stage_manager.stage)
+        held_layers.extend(module.layers[start_idx:end_idx])
+        if stage_manager.is_last_stage():
+            held_layers.append(module.norm)
+        return held_layers
+
+    def get_shared_params(self) -> List[Dict[int, Tensor]]:
+        """No shared params in bert model"""
+        return []
+
+
 class LlamaForCausalLMPolicy(LlamaPolicy):
 
     def module_policy(self):
@@ -144,3 +200,373 @@ def module_policy(self):
             }
             policy.update(new_item)
         return policy
+
+
+def llama_model_forward(
+    self: LlamaModel,
+    input_ids: torch.LongTensor = None,
+    attention_mask: Optional[torch.Tensor] = None,
+    position_ids: Optional[torch.LongTensor] = None,
+    past_key_values: Optional[List[torch.FloatTensor]] = None,
+    inputs_embeds: Optional[torch.FloatTensor] = None,
+    use_cache: Optional[bool] = None,
+    output_attentions: Optional[bool] = None,
+    output_hidden_states: Optional[bool] = None,
+    return_dict: Optional[bool] = None,
+    stage_manager: Optional[PipelineStageManager] = None,
+    hidden_states: Optional[torch.FloatTensor] = None,
+    stage_index: Optional[List[int]] = None,
+):
+    output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions
+    output_hidden_states = (output_hidden_states
+                            if output_hidden_states is not None else self.config.output_hidden_states)
+    use_cache = use_cache if use_cache is not None else self.config.use_cache
+
+    return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+
+    # retrieve input_ids and inputs_embeds
+    if stage_manager.is_first_stage():
+        if input_ids is not None and inputs_embeds is not None:
+            raise ValueError("You cannot specify both decoder_input_ids and decoder_inputs_embeds at the same time")
+        elif input_ids is not None:
+            batch_size, seq_length = input_ids.shape
+        elif inputs_embeds is not None:
+            batch_size, seq_length, _ = inputs_embeds.shape
+        else:
+            raise ValueError("You have to specify either decoder_input_ids or decoder_inputs_embeds")
+        device = input_ids.device if input_ids is not None else inputs_embeds.device
+        if inputs_embeds is None:
+            inputs_embeds = self.embed_tokens(input_ids)
+        hidden_states = inputs_embeds
+    else:
+        input_shape = hidden_states.shape[:-1]
+        batch_size, seq_length = input_shape
+        device = hidden_states.device
+
+    seq_length_with_past = seq_length
+    past_key_values_length = 0
+
+    # TODO: left the recording kv-value tensors as () or None type, this feature may be added in the future.
+    if output_attentions:
+        logger.warning_once('output_attentions=True is not supported for pipeline models at the moment.')
+        output_attentions = False
+    if output_hidden_states:
+        logger.warning_once('output_hidden_states=True is not supported for pipeline models at the moment.')
+        output_hidden_states = False
+    if use_cache:
+        logger.warning_once('use_cache=True is not supported for pipeline models at the moment.')
+        use_cache = False
+
+    if past_key_values is not None:
+        past_key_values_length = past_key_values[0][0].shape[2]
+        seq_length_with_past = seq_length_with_past + past_key_values_length
+
+    if position_ids is None:
+        position_ids = torch.arange(past_key_values_length,
+                                    seq_length + past_key_values_length,
+                                    dtype=torch.long,
+                                    device=device)
+        position_ids = position_ids.unsqueeze(0).view(-1, seq_length)
+    else:
+        position_ids = position_ids.view(-1, seq_length).long()
+
+    # embed positions, for the first stage, hidden_states is the input embeddings,
+    # for the other stages, hidden_states is the output of the previous stage
+    if attention_mask is None:
+        attention_mask = torch.ones((batch_size, seq_length_with_past), dtype=torch.bool, device=hidden_states.device)
+    attention_mask = self._prepare_decoder_attention_mask(attention_mask, (batch_size, seq_length), hidden_states,
+                                                          past_key_values_length)
+
+    if self.gradient_checkpointing and self.training:
+        if use_cache:
+            logger.warning_once(
+                "`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`...")
+            use_cache = False
+
+    # decoder layers
+    all_hidden_states = () if output_hidden_states else None
+    all_self_attns = () if output_attentions else None
+    next_decoder_cache = () if use_cache else None
+
+    start_idx, end_idx = stage_index[0], stage_index[1]
+    for idx, decoder_layer in enumerate(self.layers[start_idx:end_idx]):
+        if output_hidden_states:
+            all_hidden_states += (hidden_states,)
+
+        past_key_value = past_key_values[idx] if past_key_values is not None else None
+
+        if self.gradient_checkpointing and self.training:
+
+            def create_custom_forward(module):
+
+                def custom_forward(*inputs):
+                    # None for past_key_value
+                    return module(*inputs, output_attentions, None)
+
+                return custom_forward
+
+            layer_outputs = torch.utils.checkpoint.checkpoint(
+                create_custom_forward(decoder_layer),
+                hidden_states,
+                attention_mask,
+                position_ids,
+                None,
+            )
+        else:
+            layer_outputs = decoder_layer(
+                hidden_states,
+                attention_mask=attention_mask,
+                position_ids=position_ids,
+                past_key_value=past_key_value,
+                output_attentions=output_attentions,
+                use_cache=use_cache,
+            )
+
+        hidden_states = layer_outputs[0]
+
+        if use_cache:
+            next_decoder_cache += (layer_outputs[2 if output_attentions else 1],)
+        if output_attentions:
+            all_self_attns += (layer_outputs[1],)
+
+    if stage_manager.is_last_stage():
+        hidden_states = self.norm(hidden_states)
+
+    # add hidden states from the last decoder layer
+    if output_hidden_states:
+        all_hidden_states += (hidden_states,)
+    next_cache = next_decoder_cache if use_cache else None
+    if stage_manager.is_last_stage():
+        if not return_dict:
+            return tuple(v for v in [hidden_states, next_cache, all_hidden_states, all_self_attns] if v is not None)
+        return BaseModelOutputWithPast(
+            last_hidden_state=hidden_states,
+            past_key_values=next_cache,
+            hidden_states=all_hidden_states,
+            attentions=all_self_attns,
+        )
+    # always return dict for imediate stage
+    return {'hidden_states': hidden_states}
+
+
+def llama_for_causal_lm_forward(
+    self: LlamaForCausalLM,
+    input_ids: torch.LongTensor = None,
+    attention_mask: Optional[torch.Tensor] = None,
+    position_ids: Optional[torch.LongTensor] = None,
+    past_key_values: Optional[List[torch.FloatTensor]] = None,
+    inputs_embeds: Optional[torch.FloatTensor] = None,
+    labels: Optional[torch.LongTensor] = None,
+    use_cache: Optional[bool] = None,
+    output_attentions: Optional[bool] = None,
+    output_hidden_states: Optional[bool] = None,
+    return_dict: Optional[bool] = None,
+    stage_manager: Optional[PipelineStageManager] = None,
+    hidden_states: Optional[torch.FloatTensor] = None,
+    stage_index: Optional[List[int]] = None,
+):
+    r"""
+        Args:
+            labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
+                Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
+                config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored
+                (masked), the loss is only computed for the tokens with labels in `[0, ..., config.vocab_size]`.
+
+        Returns:
+
+        Example:
+
+        ```python
+        >>> from transformers import AutoTokenizer, LlamaForCausalLM
+
+        >>> model = LlamaForCausalLM.from_pretrained(PATH_TO_CONVERTED_WEIGHTS)
+        >>> tokenizer = AutoTokenizer.from_pretrained(PATH_TO_CONVERTED_TOKENIZER)
+
+        >>> prompt = "Hey, are you consciours? Can you talk to me?"
+        >>> inputs = tokenizer(prompt, return_tensors="pt")
+
+        >>> # Generate
+        >>> generate_ids = model.generate(inputs.input_ids, max_length=30)
+        >>> tokenizer.batch_decode(generate_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0]
+        "Hey, are you consciours? Can you talk to me?\nI'm not consciours, but I can talk to you."
+        ```"""
+    output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions
+    output_hidden_states = (output_hidden_states
+                            if output_hidden_states is not None else self.config.output_hidden_states)
+    return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+
+    # TODO: left the recording kv-value tensors as () or None type, this feature may be added in the future.
+    if output_attentions:
+        logger.warning_once('output_attentions=True is not supported for pipeline models at the moment.')
+        output_attentions = False
+    if output_hidden_states:
+        logger.warning_once('output_hidden_states=True is not supported for pipeline models at the moment.')
+        output_hidden_states = False
+    if return_dict:
+        logger.warning_once('return_dict is not supported for pipeline models at the moment')
+        return_dict = False
+
+    # decoder outputs consists of (dec_features, layer_state, dec_hidden, dec_attn)
+    outputs = llama_model_forward(
+        self.model,
+        input_ids=input_ids,
+        attention_mask=attention_mask,
+        position_ids=position_ids,
+        past_key_values=past_key_values,
+        inputs_embeds=inputs_embeds,
+        use_cache=use_cache,
+        output_attentions=output_attentions,
+        output_hidden_states=output_hidden_states,
+        return_dict=return_dict,
+        stage_manager=stage_manager,
+        hidden_states=hidden_states,
+        stage_index=stage_index,
+    )
+    past_key_values = None
+    all_hidden_states = None
+    all_self_attentions = None
+    all_cross_attentions = None
+
+    if stage_manager.is_last_stage():
+        hidden_states = outputs[0]
+        logits = self.lm_head(hidden_states)
+        loss = None
+        if labels is not None:
+            # Shift so that tokens < n predict n
+            shift_logits = logits[..., :-1, :].contiguous()
+            shift_labels = labels[..., 1:].contiguous()
+            # Flatten the tokens
+            loss_fct = CrossEntropyLoss()
+            shift_logits = shift_logits.view(-1, self.config.vocab_size)
+            shift_labels = shift_labels.view(-1)
+            # Enable model parallelism
+            shift_labels = shift_labels.to(shift_logits.device)
+            loss = loss_fct(shift_logits, shift_labels)
+
+        if not return_dict:
+            output = (logits,) + outputs[1:]
+            return (loss,) + output if loss is not None else output
+
+        return CausalLMOutputWithPast(
+            loss=loss,
+            logits=logits,
+            past_key_values=outputs.past_key_values,
+            hidden_states=outputs.hidden_states,
+            attentions=outputs.attentions,
+        )
+    else:
+        hidden_states = outputs.get('hidden_states')
+        return {'hidden_states': hidden_states}
+
+
+def llama_for_sequence_classification_forward(
+    self: LlamaForSequenceClassification,
+    input_ids: torch.LongTensor = None,
+    attention_mask: Optional[torch.Tensor] = None,
+    position_ids: Optional[torch.LongTensor] = None,
+    past_key_values: Optional[List[torch.FloatTensor]] = None,
+    inputs_embeds: Optional[torch.FloatTensor] = None,
+    labels: Optional[torch.LongTensor] = None,
+    use_cache: Optional[bool] = None,
+    output_attentions: Optional[bool] = None,
+    output_hidden_states: Optional[bool] = None,
+    return_dict: Optional[bool] = None,
+    stage_manager: Optional[PipelineStageManager] = None,
+    hidden_states: Optional[torch.FloatTensor] = None,
+    stage_index: Optional[List[int]] = None,
+):
+    r"""
+    labels (`torch.LongTensor` of shape `(batch_size,)`, *optional*):
+        Labels for computing the sequence classification/regression loss. Indices should be in `[0, ...,
+        config.num_labels - 1]`. If `config.num_labels == 1` a regression loss is computed (Mean-Square loss), If
+        `config.num_labels > 1` a classification loss is computed (Cross-Entropy).
+    """
+    return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+    # TODO: left the recording kv-value tensors as () or None type, this feature may be added in the future.
+    if output_attentions:
+        logger.warning_once('output_attentions=True is not supported for pipeline models at the moment.')
+        output_attentions = False
+    if output_hidden_states:
+        logger.warning_once('output_hidden_states=True is not supported for pipeline models at the moment.')
+        output_hidden_states = False
+    if return_dict:
+        logger.warning_once('return_dict is not supported for pipeline models at the moment')
+        return_dict = False
+
+    transformer_outputs = llama_model_forward(
+        self.model,
+        input_ids,
+        attention_mask=attention_mask,
+        position_ids=position_ids,
+        past_key_values=past_key_values,
+        inputs_embeds=inputs_embeds,
+        use_cache=use_cache,
+        output_attentions=output_attentions,
+        output_hidden_states=output_hidden_states,
+        return_dict=return_dict,
+        stage_manager=stage_manager,
+        hidden_states=hidden_states,
+        stage_index=stage_index,
+    )
+
+    if input_ids is not None:
+        batch_size = input_ids.shape[0]
+    elif inputs_embeds is not None:
+        batch_size = inputs_embeds.shape[0]
+    else:
+        batch_size = hidden_states.shape[0]
+
+    if stage_manager.is_last_stage():
+        hidden_states = transformer_outputs[0]
+        logits = self.score(hidden_states)
+
+        if self.config.pad_token_id is None and batch_size != 1:
+            raise ValueError("Cannot handle batch sizes > 1 if no padding token is defined.")
+        if self.config.pad_token_id is None:
+            sequence_lengths = -1
+        else:
+            if input_ids is not None:
+                sequence_lengths = (torch.ne(input_ids, self.config.pad_token_id).sum(-1) - 1).to(logits.device)
+            else:
+                sequence_lengths = -1
+
+        pooled_logits = logits[torch.arange(batch_size, device=logits.device), sequence_lengths]
+
+        loss = None
+        if labels is not None:
+            labels = labels.to(logits.device)
+            if self.config.problem_type is None:
+                if self.num_labels == 1:
+                    self.config.problem_type = "regression"
+                elif self.num_labels > 1 and (labels.dtype == torch.long or labels.dtype == torch.int):
+                    self.config.problem_type = "single_label_classification"
+                else:
+                    self.config.problem_type = "multi_label_classification"
+
+            if self.config.problem_type == "regression":
+                loss_fct = MSELoss()
+                if self.num_labels == 1:
+                    loss = loss_fct(pooled_logits.squeeze(), labels.squeeze())
+                else:
+                    loss = loss_fct(pooled_logits, labels)
+            elif self.config.problem_type == "single_label_classification":
+                loss_fct = CrossEntropyLoss()
+                loss = loss_fct(pooled_logits.view(-1, self.num_labels), labels.view(-1))
+            elif self.config.problem_type == "multi_label_classification":
+                loss_fct = BCEWithLogitsLoss()
+                loss = loss_fct(pooled_logits, labels)
+        if not return_dict:
+            output = (pooled_logits,) + transformer_outputs[1:]
+            return ((loss,) + output) if loss is not None else output
+
+        return SequenceClassifierOutputWithPast(
+            loss=loss,
+            logits=pooled_logits,
+            past_key_values=transformer_outputs.past_key_values,
+            hidden_states=transformer_outputs.hidden_states,
+            attentions=transformer_outputs.attentions,
+        )
+
+    else:
+        hidden_states = transformer_outputs.get('hidden_states')
+        return {'hidden_states': hidden_states}
diff --git a/tests/kit/model_zoo/transformers/gpt.py b/tests/kit/model_zoo/transformers/gpt.py
index b9e0310780af..ac70138e3f8f 100644
--- a/tests/kit/model_zoo/transformers/gpt.py
+++ b/tests/kit/model_zoo/transformers/gpt.py
@@ -52,7 +52,7 @@ def data_gen_for_sequence_classification():
 loss_fn = lambda x: x.loss
 
 config = transformers.GPT2Config(n_layer=2,
-                                 n_head=4,
+                                 n_head=2,
                                  vocab_size=50258,
                                  attn_pdrop=0,
                                  embd_pdrop=0,
diff --git a/tests/test_shardformer/test_model/_utils.py b/tests/test_shardformer/test_model/_utils.py
index de8cb65d21d0..f26c6622da7e 100644
--- a/tests/test_shardformer/test_model/_utils.py
+++ b/tests/test_shardformer/test_model/_utils.py
@@ -39,6 +39,7 @@ def build_pipeline_model(model_fn,
     shard_config = ShardConfig(enable_fused_normalization=enable_fused_normalization,
                                enable_tensor_parallelism=enable_tensor_parallelism,
                                pipeline_stage_manager=stage_manager)
+
     shard_former = ShardFormer(shard_config=shard_config)
     sharded_model, shared_params = shard_former.optimize(model_copy)
     return org_model.cuda(), sharded_model.cuda()
diff --git a/tests/test_shardformer/test_model/test_shard_llama_pipeline.py b/tests/test_shardformer/test_model/test_shard_llama_pipeline.py
new file mode 100644
index 000000000000..81c183d3230e
--- /dev/null
+++ b/tests/test_shardformer/test_model/test_shard_llama_pipeline.py
@@ -0,0 +1,85 @@
+import pytest
+import torch
+
+import colossalai
+from colossalai.cluster import ProcessGroupMesh
+from colossalai.logging import disable_existing_loggers
+from colossalai.pipeline.stage_manager import PipelineStageManager
+from colossalai.shardformer.policies.base_policy import Policy
+from colossalai.tensor.d_tensor.api import is_customized_distributed_tensor, is_distributed_tensor
+from colossalai.testing import (
+    assert_hf_output_close,
+    clear_cache_before_run,
+    parameterize,
+    rerun_if_address_is_in_use,
+    spawn,
+)
+from tests.kit.model_zoo import model_zoo
+from tests.test_shardformer.test_model._utils import build_model, build_pipeline_model, run_forward
+
+
+def check_forward_backward(org_model, sharded_model, data_gen_fn, output_transform_fn, loss_fn):
+    # check forward
+    pass
+
+
+@parameterize('enable_fused_normalization', [False])
+@parameterize('enable_tensor_parallelism', [False])
+@parameterize('use_lazy_init', [False])
+#TODO: merge this into test_shard_llama
+def run_llama_test(enable_fused_normalization, enable_tensor_parallelism, use_lazy_init):
+    DP_DIM, PP_DIM = 0, 1
+    DP_SIZE, PP_SIZE = 2, 2
+    RANK_TO_COORDINATE = {
+        0: (0, 0),
+        1: (0, 1),
+        2: (1, 0),
+        3: (1, 1),
+    }
+    PP_RANKS_IN_GROUP = {
+        0: [0, 1],
+        1: [0, 1],
+        2: [2, 3],
+        3: [2, 3],
+    }
+    pg_mesh = ProcessGroupMesh(DP_SIZE, PP_SIZE)
+    stage_manager = PipelineStageManager(pg_mesh, PP_DIM)
+
+    sub_model_zoo = model_zoo.get_sub_registry('transformers_llama')
+    x = torch.randint(0, 1000, (2, 3)).cuda()
+    hidden_states = torch.randint(0, 1000, (2, 3, 128)).to(torch.float32).cuda()
+    for name, (model_fn, data_gen_fn, output_transform_fn, loss_fn, _) in sub_model_zoo.items():
+        if name == 'transformers_llama':
+            org_model, sharded_model = build_pipeline_model(model_fn, stage_manager, enable_fused_normalization,
+                                                            enable_tensor_parallelism, use_lazy_init)
+            if stage_manager.stage == 0:
+                attention_mask = torch.ones_like(x).cuda()
+                output = sharded_model(input_ids=x, attention_mask=attention_mask)
+                assert output['hidden_states'].shape == (2, 3, 128)
+            else:
+                attention_mask = torch.ones((2, 3)).cuda()
+                output = sharded_model(
+                    hidden_states=hidden_states,
+                    attention_mask=attention_mask,
+                )
+                # print(output[0].shape)
+                assert output[0].shape == (2, 3, 128)
+
+    torch.cuda.empty_cache()
+
+
+def check_llama(rank, world_size, port):
+    disable_existing_loggers()
+    colossalai.launch(config={}, rank=rank, world_size=world_size, host='localhost', port=port, backend='nccl')
+    run_llama_test()
+
+
+@pytest.mark.dist
+@rerun_if_address_is_in_use()
+@clear_cache_before_run()
+def test_llama():
+    spawn(check_llama, 4)
+
+
+if __name__ == "__main__":
+    test_llama()