databricks-ml-examples
Example notebooks for the mistral 7B model on Databricks
This folder contains the following examples for mistral-7b models: ` <!—
–>
| File | Description | Model Used | GPU Minimum Requirement |
|---|---|---|---|
| 01_load_inference | Environment setup and suggested configurations when inferencing mistral-7b-instruct models on Databricks. | mistral-7b-instruct |
1xA10-24GB |
| 02_mlflow_logging_inference | Save, register, and load mistral-7b-instruct models with MLflow, and create a Databricks model serving endpoint. | mistral-7b-instruct |
1xA10-24GB |
| [02[chat]_mlflow_logging_inference](https://github.com/databricks/databricks-ml-examples/blob/master/llm-models/mistral/mistral-7b/02[chat]_mlflow_logging_inference.py) | Save, register, and load mistral-7b-instruct models with MLflow, and create a Databricks model serving endpoint for chat completion. | mistral-7b-instruct |
1xA10-24GB |
| 03_serve_driver_proxy | Serve mistral-7b-instruct models on the cluster driver node using Flask. | mistral-7b-instruct |
1xA10-24GB |
| [03[chat]_serve_driver_proxy](https://github.com/databricks/databricks-ml-examples/blob/master/llm-models/mistral/mistral-7b/03[chat]_serve_driver_proxy.py) | Serve mistral-7b-instruct models as chat completion on the cluster driver node using Flask. | mistral-7b-instruct |
1xA10-24GB |
| 04_langchain | Integrate a serving endpoint or cluster driver proxy app with LangChain and query. | N/A | N/A |
| [04[chat]_langchain](https://github.com/databricks/databricks-ml-examples/blob/master/llm-models/mistral/mistral-7b/04[chat]_langchain.py) | Integrate a serving endpoint and setup langchain chat model. | N/A | N/A |
| 05_fine_tune_deepspeed | Fine-tune mistral-7b models leveraging DeepSpeed. | mistral-7b |
4xA10 or 2xA100-80GB |
| 06_fine_tune_qlora | Fine-tune mistral-7b models with QLORA. | mistral-7b |
1xA10 |
| 07_ai_gateway | Manage a MLflow AI Gateway Route that accesses a Databricks model serving endpoint. | N/A | N/A |