diff --git a/torchbenchmark/models/hf_MPT_7b_instruct/__init__.py b/torchbenchmark/models/hf_MPT_7b_instruct/__init__.py
new file mode 100644
index 0000000000..59f314bdd3
--- /dev/null
+++ b/torchbenchmark/models/hf_MPT_7b_instruct/__init__.py
@@ -0,0 +1,14 @@
+from torchbenchmark.tasks import NLP
+from torchbenchmark.util.framework.huggingface.model_factory import HuggingFaceModel
+
+class Model(HuggingFaceModel):
+    task = NLP.LANGUAGE_MODELING
+    # https://huggingface.co/mosaicml/mpt-7b
+    DEFAULT_TRAIN_BSIZE = 4
+    DEFAULT_EVAL_BSIZE = 1
+
+    def __init__(self, test, device, batch_size=None, extra_args=[]):
+        super().__init__(name="hf_MPT_7b_instruct", test=test, device=device, batch_size=batch_size, extra_args=extra_args)
+
+    def eval(self):
+        super().eval()
\ No newline at end of file
diff --git a/torchbenchmark/models/hf_MPT_7b_instruct/install.py b/torchbenchmark/models/hf_MPT_7b_instruct/install.py
new file mode 100644
index 0000000000..152b955c4c
--- /dev/null
+++ b/torchbenchmark/models/hf_MPT_7b_instruct/install.py
@@ -0,0 +1,7 @@
+import os
+from torchbenchmark.util.framework.huggingface.patch_hf import patch_transformers, cache_model
+
+if __name__ == '__main__':
+    patch_transformers()
+    model_name = os.path.basename(os.path.dirname(os.path.abspath(__file__)))
+    cache_model(model_name, trust_remote_code=True)
\ No newline at end of file
diff --git a/torchbenchmark/models/hf_MPT_7b_instruct/metadata.yaml b/torchbenchmark/models/hf_MPT_7b_instruct/metadata.yaml
new file mode 100644
index 0000000000..c525275e5d
--- /dev/null
+++ b/torchbenchmark/models/hf_MPT_7b_instruct/metadata.yaml
@@ -0,0 +1,8 @@
+devices:
+  NVIDIA A100-SXM4-40GB:
+    eval_batch_size: 1
+eval_benchmark: false
+eval_deterministic: false
+eval_nograd: true
+train_benchmark: false
+train_deterministic: false
\ No newline at end of file
diff --git a/torchbenchmark/util/framework/huggingface/model_factory.py b/torchbenchmark/util/framework/huggingface/model_factory.py
index d317483a67..2f4c202078 100644
--- a/torchbenchmark/util/framework/huggingface/model_factory.py
+++ b/torchbenchmark/util/framework/huggingface/model_factory.py
@@ -30,6 +30,7 @@
     'hf_Whisper': (1024, 1024, 'WhisperConfig()', 'AutoModelForAudioClassification'),
     # default num_hidden_layers=32 but that OOMs, feel free to change this config to something more real
     'llama_v2_7b_16h' : (512,512, 'LlamaConfig(num_hidden_layers=16)', 'AutoModelForCausalLM'),
+    'hf_MPT_7b_instruct': (512, 512, 'AutoConfig.from_pretrained("mosaicml/mpt-7b-instruct", trust_remote_code=True)', 'AutoModelForCausalLM'),
     'llama_v2_7b' : (512,512, 'AutoConfig.from_pretrained("meta-llama/Llama-2-7b-hf")', 'AutoModelForCausalLM'),
     'llama_v2_13b' : (512,512, 'AutoConfig.from_pretrained("meta-llama/Llama-2-13b-hf")', 'AutoModelForCausalLM'),
     'llama_v2_70b' : (512, 512, 'AutoConfig.from_pretrained("meta-llama/Llama-2-70b-hf")', 'AutoModelForMaskedLM'),
@@ -83,7 +84,10 @@ def __init__(self, name, test, device, batch_size=None, extra_args=[]):
             # silence "config.num_buckets is not set. Setting config.num_buckets to 128"
             config.num_buckets = 128
         class_ctor = getattr(transformers, class_models[name][3])
-        self.model = class_ctor.from_config(config).to(device)
+        kwargs = {}
+        if name == "hf_Falcon_7b" or name == "hf_MPT_7b_instruct":
+            kwargs["trust_remote_code"] = True
+        self.model = class_ctor.from_config(config, **kwargs).to(device)
         self.optimizer = optim.Adam(
             self.model.parameters(),
             lr=0.001,
diff --git a/torchbenchmark/util/framework/huggingface/patch_hf.py b/torchbenchmark/util/framework/huggingface/patch_hf.py
index 24de3f1eb3..f23055be64 100644
--- a/torchbenchmark/util/framework/huggingface/patch_hf.py
+++ b/torchbenchmark/util/framework/huggingface/patch_hf.py
@@ -10,11 +10,11 @@
 
 PATCH_DIR = os.path.join(os.path.dirname(os.path.abspath(__file__)), "patches")
 
-def cache_model(name: str):
+def cache_model(name: str, **kwargs):
     import transformers
     model_config = eval(class_models[name][2])
     model_ctor = getattr(transformers, class_models[name][3])
-    model_ctor.from_config(model_config)
+    model_ctor.from_config(model_config, **kwargs)
 
 def patch_transformers():
     import transformers