File size: 2,915 Bytes
db25206 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 |
---
base_model: TinyLlama/TinyLlama-1.1B-Chat-v1.0
inference: false
model_type: llama
prompt_template: |
<|im_start|>user\n
{prompt}<|im_end|>\n
<|im_start|>assistant\n
quantized_by: mwitiderrick
tags:
- deepsparse
---
## TinyLlama 1.1B Chat 0.4 - DeepSparse
This repo contains model files for [TinyLlama 1.1B Chat](https://huggingface.co/TinyLlama/TinyLlama-1.1B-Chat-v1.0) optimized for [DeepSparse](https://github.com/neuralmagic/deepsparse), a CPU inference runtime for sparse models.
This model was quantized and pruned with [SparseGPT](https://arxiv.org/abs/2301.00774), using [SparseML](https://github.com/neuralmagic/sparseml).
## Inference
Install [DeepSparse LLM](https://github.com/neuralmagic/deepsparse) for fast inference on CPUs:
```bash
pip install deepsparse-nightly[llm]
```
Run in a [Python pipeline](https://github.com/neuralmagic/deepsparse/blob/main/docs/llms/text-generation-pipeline.md):
```python
from deepsparse import TextGeneration
prompt = "How to make banana bread?"
formatted_prompt = f"<|im_start|>user\n{prompt}<|im_end|>\n<|im_start|>assistant\n"
model = TextGeneration(model="hf:neuralmagic/TinyLlama-1.1B-Chat-v0.4-pruned50-quant-ds")
print(model(formatted_prompt, max_new_tokens=500).generations[0].text)
"""
"""
```
## Prompt template
```
<|im_start|>user\n
{prompt}<|im_end|>\n
<|im_start|>assistant\n
```
## Sparsification
For details on how this model was sparsified, see the `recipe.yaml` in this repo and follow the instructions below.
```bash
git clone https://github.com/neuralmagic/sparseml
pip install -e "sparseml[transformers]"
python sparseml/src/sparseml/transformers/sparsification/obcq/obcq.py TinyLlama/TinyLlama-1.1B-Chat-v1.0 open_platypus --precision float16 --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:
```python
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](https://github.com/neuralmagic/sparseml/tree/main/src/sparseml/transformers/sparsification/obcq) 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](https://join.slack.com/t/discuss-neuralmagic/shared_invite/zt-q1a1cnvo-YBoICSIw3L1dmQpjBeDurQ) |