Instructions to use nvidia/Llama-3_3-Nemotron-Super-49B-v1_5 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use nvidia/Llama-3_3-Nemotron-Super-49B-v1_5 with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="nvidia/Llama-3_3-Nemotron-Super-49B-v1_5", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("nvidia/Llama-3_3-Nemotron-Super-49B-v1_5", trust_remote_code=True, dtype="auto")

Inference
Local Apps Settings

vLLM

How to use nvidia/Llama-3_3-Nemotron-Super-49B-v1_5 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "nvidia/Llama-3_3-Nemotron-Super-49B-v1_5"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "nvidia/Llama-3_3-Nemotron-Super-49B-v1_5",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/nvidia/Llama-3_3-Nemotron-Super-49B-v1_5

SGLang

How to use nvidia/Llama-3_3-Nemotron-Super-49B-v1_5 with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "nvidia/Llama-3_3-Nemotron-Super-49B-v1_5" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "nvidia/Llama-3_3-Nemotron-Super-49B-v1_5",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "nvidia/Llama-3_3-Nemotron-Super-49B-v1_5" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "nvidia/Llama-3_3-Nemotron-Super-49B-v1_5",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use nvidia/Llama-3_3-Nemotron-Super-49B-v1_5 with Docker Model Runner:
```
docker model run hf.co/nvidia/Llama-3_3-Nemotron-Super-49B-v1_5
```

NePe commited on Nov 29, 2025

Commit

2be40ac

verified ·

1 Parent(s): 36f62ed

Remove incompatible code + method to get VariableCache

Browse files

Usage:
past_key_values = model.getVariableCache(batch_size=1, max_cache_len=4096)
model.generate(... ,past_key_values=past_key_values)

Files changed (1) hide show

modeling_decilm.py +7 -3

modeling_decilm.py CHANGED Viewed

@@ -27,7 +27,7 @@ import torch.utils.checkpoint
 from torch import nn
 from torch.nn import BCEWithLogitsLoss, CrossEntropyLoss, MSELoss
 from transformers import GenerationConfig
-from transformers.generation.utils import NEED_SETUP_CACHE_CLASSES_MAPPING, GenerationMixin, GenerateOutput
 from transformers.modeling_utils import PreTrainedModel
 from transformers.models.auto.modeling_auto import MODEL_FOR_CAUSAL_LM_MAPPING_NAMES
 from transformers.utils import (
@@ -809,8 +809,9 @@ class DeciLMPreTrainedModel(PreTrainedModel):
     ) -> tuple[GenerationConfig, dict]:
         # DeciLM-specific code
         generation_config, model_kwargs = super()._prepare_generation_config(generation_config, *args, **kwargs)
-        generation_config.cache_implementation = "variable"
-        NEED_SETUP_CACHE_CLASSES_MAPPING["variable"] = VariableCache
         return generation_config, model_kwargs
@@ -1133,6 +1134,9 @@ class DeciLMForCausalLM(DeciLMPreTrainedModel, GenerationMixin):
     def get_decoder(self):
         return self.model
     @add_start_docstrings_to_model_forward(DECILM_INPUTS_DOCSTRING)
     @replace_return_docstrings(output_type=CausalLMOutputWithPast, config_class=_CONFIG_FOR_DOC)
     def forward(

 from torch import nn
 from torch.nn import BCEWithLogitsLoss, CrossEntropyLoss, MSELoss
 from transformers import GenerationConfig
+from transformers.generation.utils import GenerationMixin, GenerateOutput
 from transformers.modeling_utils import PreTrainedModel
 from transformers.models.auto.modeling_auto import MODEL_FOR_CAUSAL_LM_MAPPING_NAMES
 from transformers.utils import (
     ) -> tuple[GenerationConfig, dict]:
         # DeciLM-specific code
         generation_config, model_kwargs = super()._prepare_generation_config(generation_config, *args, **kwargs)
+        generation_config.disable_compile = True
+        #generation_config.cache_implementation = "variable"
+        #NEED_SETUP_CACHE_CLASSES_MAPPING["variable"] = VariableCache
         return generation_config, model_kwargs
     def get_decoder(self):
         return self.model
+    def getVariableCache(self, batch_size=1, max_cache_len=4096, dtype=torch.bfloat16):
+        return VariableCache(config=self.config, batch_size=batch_size, max_batch_size=batch_size, max_cache_len=max_cache_len, dtype=dtype)
     @add_start_docstrings_to_model_forward(DECILM_INPUTS_DOCSTRING)
     @replace_return_docstrings(output_type=CausalLMOutputWithPast, config_class=_CONFIG_FOR_DOC)
     def forward(