This repository does the following:
- Fine tuning an LLM
- Populate a vector database with and embedding model, so able to query your context similarty in the vector database
- Fine tune with Ray framework
- Use CPU and GPU for fine tuning and serving
- Serve your fine tuned LLM as REST API.
Please set the API keys accordingly save the content below in llm_agent/.env
.
OPENAI_API_KEY=
ANYSCALE_API_KEY=
OPENAI_API_BASE="https://api.endpoints.anyscale.com/v1"
ANYSCALE_API_BASE="https://api.endpoints.anyscale.com/v1"
DB_CONNECTION_STRING="postgresql://testUser:testPassword@localhost:15432/testDB"
EMBEDDING_INDEX_DIR=/tmp/embedding_index_sql
VECTOR_TABLE_NAME=document
VECTOR_TABLE_DUMP_OUTPUT_PATH=/tmp/vector.document.dump.sql
RAYDOCS_ROOT=/tmp/raydocs
NUM_CPUS=14
NUM_GPUS=1
NUM_CHUNKS=5
CHUNK_SIZE=500
CHUNK_OVERLAP=50
EMBEDDING_MODEL_NAME="thenlper/gte-base"
LLM_MODEL_NAME=meta-llama/Llama-2-70b-chat-hf
#How much data should be fed for fine tuning
#give a floating number between >0.001 and 1 (1 included, which means use all the data for fine tuning)
USE_THIS_PORTION_OF_DATA=0.05
make scrape # Scrap the web page
make vectordb # Configure Postgres Vector DB
make postgres-client # Install Postgres Client
Then, in a seperate terminal
make port-forward-postgres # Port Forward DB
make vector-support # Enable Vector Support
make vector-table # Create Vector Table
make embedding-table # Get Vector Table
# result:
List of relations
Schema | Name | Type | Owner
--------+-----------------+----------+----------
public | document | table | testUser
public | document_id_seq | sequence | testUser
(2 rows)
make pods-preview # Get Pods
make install-pip-deps # Install Pip Dependencies
Once Setup, the following commands enable finetuning on a ray cluster:
make ray-cluster # Start Ray Cluster
make profile-ray-cluster # Profile Cluster
make finetune # Finetune LLM
At the end you will see something like below:
The default batch size for map_batches is rollout_fragment_length * num_envs.
which indicates that LLM fine tuning is done, vector db is populated, and a query is sent to LLM with the context identified by your vector DB.
Note: My machine has 16 CPUs and 1 GPU, so I set up NUM_CPUS
and NUM_GPUs
accordingly. These numbers may differ according to your machine. The principle here is that you can not set up a number larger than existing resources (CPU and GPU).
Pleae note that we are using thenlper/gte-base
as an embedding model, this is a relatively small model, you might like to change it. LLM_MODEL_NAME
is to meta-llama/Llama-2-70b-chat-hf
, which is good for this setup, but again you might like to change it.
make dev-deploy
make test-query
Should yield:
b'"{\\"question\\": \\"What is the default batch size for map_batches?\\", \\"sources\\": [\\"https://docs.ray.io/en/master/rllib/rllib-training.html#specifying-rollout-workers\\", \\"https://docs.ray.io/en/master/rllib/rllib-training.html#specifying-rollout-workers\\", \\"https://docs.ray.io/en/master/rllib/package_ref/doc/ray.rllib.policy.policy.Policy.compute_log_likelihoods.html#ray-rllib-policy-policy-policy-compute-log-likelihoods\\", \\"https://docs.ray.io/en/master/rllib/package_ref/doc/ray.rllib.policy.policy.Policy.compute_log_likelihoods.html#ray-rllib-policy-policy-policy-compute-log-likelihoods\\", \\"https://docs.ray.io/en/master/rllib/rllib-algorithms.html#importance-weighted-actor-learner-architecture-impala\\"], \\"answer\\": \\" The default batch size for map_batches is rollout_fragment_length * num_envs.\\", \\"llm\\": \\"meta-llama/Llama-2-70b-chat-hf\\"}"'
- Spot Instance/Fleet Provisioning for Cost Effective Training
- CUDA devcontainer configurations
- Dockerfiles
- Terraform Configuration for 3-Tier Cloud Deployment
- linting, testing
- Github Push/Pull Actions + CI/CD Building
- Intergrating Other DB Backends
- Quantization
1 A Comprehensive Guide for Building RAG-based LLM Applications (Part 1). Any Scale Blog.