Instructions to use newmindai/Llama-3.1-8B-Instruct-w16a8-rw-with-gw-hp-4nodes with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use newmindai/Llama-3.1-8B-Instruct-w16a8-rw-with-gw-hp-4nodes with Transformers:

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

pipe = pipeline("text-generation", model="newmindai/Llama-3.1-8B-Instruct-w16a8-rw-with-gw-hp-4nodes")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("newmindai/Llama-3.1-8B-Instruct-w16a8-rw-with-gw-hp-4nodes")
model = AutoModelForCausalLM.from_pretrained("newmindai/Llama-3.1-8B-Instruct-w16a8-rw-with-gw-hp-4nodes", device_map="auto")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use newmindai/Llama-3.1-8B-Instruct-w16a8-rw-with-gw-hp-4nodes with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "newmindai/Llama-3.1-8B-Instruct-w16a8-rw-with-gw-hp-4nodes"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "newmindai/Llama-3.1-8B-Instruct-w16a8-rw-with-gw-hp-4nodes",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/newmindai/Llama-3.1-8B-Instruct-w16a8-rw-with-gw-hp-4nodes

SGLang

How to use newmindai/Llama-3.1-8B-Instruct-w16a8-rw-with-gw-hp-4nodes 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 "newmindai/Llama-3.1-8B-Instruct-w16a8-rw-with-gw-hp-4nodes" \
    --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": "newmindai/Llama-3.1-8B-Instruct-w16a8-rw-with-gw-hp-4nodes",
		"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 "newmindai/Llama-3.1-8B-Instruct-w16a8-rw-with-gw-hp-4nodes" \
        --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": "newmindai/Llama-3.1-8B-Instruct-w16a8-rw-with-gw-hp-4nodes",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use newmindai/Llama-3.1-8B-Instruct-w16a8-rw-with-gw-hp-4nodes with Docker Model Runner:
```
docker model run hf.co/newmindai/Llama-3.1-8B-Instruct-w16a8-rw-with-gw-hp-4nodes
```

nmmursit commited on Dec 7, 2025

Commit

ce487f8

verified ·

1 Parent(s): 86cca4a

Upload README

Browse files

Files changed (1) hide show

README.md +38 -0

README.md ADDED Viewed

	@@ -0,0 +1,38 @@

+# Finetuned Model (Distributed Checkpoint)
+Base Model: meta-llama/Llama-3.1-8B-Instruct
+Dataset: newmindai/euro_hpc-legal
+Training dtype: bf16
+Checkpoint Type: Distributed/Sharded (requires loading with FSDP)
+## Loading Instructions
+This model was saved using PyTorch Distributed Checkpointing and requires special loading:
+```python
+import torch
+from torch.distributed.checkpoint import load
+from transformers import AutoModelForCausalLM, AutoTokenizer
+# Load tokenizer
+tokenizer = AutoTokenizer.from_pretrained("/gpfs/scratch/ehpc317/esavirdi/dual_precision_training/7Dec/lama3.1_8b_rowwise_with_gw_hp_4nodes/final_model")
+# Load model architecture
+model = AutoModelForCausalLM.from_pretrained(
+    "meta-llama/Llama-3.1-8B-Instruct",
+    torch_dtype=torch.bf16,
+)
+# Load sharded weights
+checkpoint = {"model": model.state_dict()}
+load(checkpoint, checkpoint_id="/gpfs/scratch/ehpc317/esavirdi/dual_precision_training/7Dec/lama3.1_8b_rowwise_with_gw_hp_4nodes/final_model/sharded")
+model.load_state_dict(checkpoint["model"])
+# Generate text
+inputs = tokenizer("Your prompt here", return_tensors="pt")
+outputs = model.generate(**inputs, max_length=100)
+print(tokenizer.decode(outputs[0]))
+```
+Note: For easier inference, consider converting this to a standard HuggingFace checkpoint
+by loading and re-saving on a single GPU.