Open-llama-3b-v2-instruct - DeepSparse
This repo contains model files for Open-llama-3b-v2-instruct optimized for DeepSparse, a CPU inference runtime for sparse models.
This model was quantized and pruned with SparseGPT, using SparseML.
Inference
Install DeepSparse LLM for fast inference on CPUs:
pip install deepsparse-nightly[llm]
Run in a Python pipeline:
from deepsparse import TextGeneration
system_message = 'You are a helpful assistant, who always provide explanation.'
user_message = 'How many days are there in a leap year?'
formatted_prompt = f'### System:\n{system_message}<|endoftext|>\n### User:\n{user_message}<|endoftext|>\n### Assistant:\n'
model = TextGeneration(model_path="hf:nm-testing/open-llama-3b-v2-instruct-pruned50-quant-ds")
print(model(formatted_prompt, max_new_tokens=200,repetition_penalty=1.05).generations[0].text)
"""
A leap year is a year that is exactly divisible by 4. So, there are 365 days in a leap year.
"""
from deepsparse import TextGeneration
user_message = 'How to make banana bread?'
prompt = f'### User:\n{user_message}\n### Assistant:\n'
model = TextGeneration(model_path="hf:nm-testing/open-llama-3b-v2-instruct-pruned50-ds")
print(model(prompt, max_new_tokens=500).generations[0].text)
"""
To make banana bread, you will need the following ingredients and steps:
Ingredients:
1. 1 1/2 cups (300 grams) of mashed bananas
2. 1 1/2 cups (300 grams) of butter
3. 1 1/2 cups (300 grams) of sugar
4. 1 1/2 cups (300 grams) of flour
5. 1 1/2 cups (300 grams) of baking powder
6. 1 1/2 cups (300 grams) of salt
Steps:
1. Preheat oven to 350°F (176°C).
2. In a large bowl, combine mashed bananas, butter, sugar, flour, baking powder, and salt.
3. Stir the mixture until it's smooth.
4. Pour the mixture into a greased baking pan.
5. Bake in the oven for 1 hour.
6. Allow the bread to cool.
7. Serve the bread.
Remember to always check the baking time and oven temperature before starting the process.
"""
Prompt template
### User:\n
{prompt}
### Assistant:\n
Sparsification
For details on how this model was sparsified, see the recipe.yaml
in this repo and follow the instructions below.
git clone https://github.com/neuralmagic/sparseml
pip install -e "sparseml[transformers]"
python sparseml/src/sparseml/transformers/sparsification/obcq/obcq.py mediocredev/open-llama-3b-v2-instruct open_platypus --recipe recipe.yaml --save True
python sparseml/src/sparseml/transformers/sparsification/obcq/export.py --task text-generation --model_path obcq_deployment
cp deployment/model.onnx deployment/model-orig.onnx
Run this kv-cache injection to speed up the model at inference by caching the Key and Value states:
import os
import onnx
from sparseml.exporters.kv_cache_injector import KeyValueCacheInjector
input_file = "deployment/model-orig.onnx"
output_file = "deployment/model.onnx"
model = onnx.load(input_file, load_external_data=False)
model = KeyValueCacheInjector(model_path=os.path.dirname(input_file)).apply(model)
onnx.save(model, output_file)
print(f"Modified model saved to: {output_file}")
Follow the instructions on our One Shot With SparseML page for a step-by-step guide for performing one-shot quantization of large language models.
Slack
For further support, and discussions on these models and AI in general, join Neural Magic's Slack Community
- Downloads last month
- 4
Inference Providers
NEW
This model is not currently available via any of the supported third-party Inference Providers, and
HF Inference API has been turned off for this model.
Model tree for nm-testing/open-llama-3b-v2-instruct-pruned50-ds
Base model
mediocredev/open-llama-3b-v2-instruct