The open source toolkit for finetuning and deploying LLMs
Project description
Arcee
The open source toolkit for finetuning and deploying LLMs
Finetune LLMS
from arcee.models import LM
from arcee.data import Instuctions
lm = LM("falcon30b")
instructions = Instructions("./datasets/stripe-api.json")
lm.train(instructions)
lm.predict("Place an order for the LLM-9000 product for 100 USD to the card 3007200039992000")
Deploy LLMS
Authenticate
import arcee
arcee.login()
Deploy to the Arcee cloud
project = arcee.create_project("stripe-api-operator")
#project = arcee.load_project(...)
#regulated under 7b params for free
#only PEFT uploadable for free
hosted_lm = project.deploy(llm)
hosted_lm.url
hosted_lm.predict("Place an order for the LLM-9000 product for 100 USD to the card 3007200039992000")
#view a streaming web app of the llm
project.demo()
LangChain Integration
from langchain import Arcee
#goes in llms/arcee.py
prompt_template = "Write a stripe API request for the following: {order}."
llm = Arcee(temperature=0)
llm_chain = LLMChain(
llm=llm,
prompt=PromptTemplate.from_template(prompt_template)
)
llm_chain("Place an order for the LLM-9000 product for 100 USD to the card 3007200039992000")
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
arcee-align-0.0.1.tar.gz
(6.3 kB
view details)
Built Distribution
File details
Details for the file arcee-align-0.0.1.tar.gz
.
File metadata
- Download URL: arcee-align-0.0.1.tar.gz
- Upload date:
- Size: 6.3 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.8.10
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 98553e4d814902a0f5678a3a6470f5d7faf5061bb17bf4361325c307abe675b3 |
|
MD5 | a6d025f9406daabc301266baa29eb778 |
|
BLAKE2b-256 | 5ed17de32ee57a28609c1b3c99c626da40c3ca854f85d0d2700e11404767f154 |
File details
Details for the file arcee_align-0.0.1-py3-none-any.whl
.
File metadata
- Download URL: arcee_align-0.0.1-py3-none-any.whl
- Upload date:
- Size: 6.6 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.8.10
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 26cb63f3bd5a5b1641c4be5594480b60a5b038328d1e73ba8869ce069fda5c23 |
|
MD5 | fced1d07dcbf4698e262ce8b026a07ab |
|
BLAKE2b-256 | 165f47897b92e2a6dc3144bd78d417466637c7b3998cfe1d5449ada8159495c5 |