This is a (slightly) faster version of KillianLucas/tokentrim for longer message arrays.
In average, gpt-trim is ~80% faster than tokentrim, and that tokentrim is around 5x~7x slower.
Although gpt-trim is fast, I still need to finish my LeetCode problems that I left years ago, just so that I can make it 20x faster than 95% of people.
The usage is quite similiar to tokentrim
.
import gpt_trim
trimmed = gpt_trim.trim(
messages,
model="gpt-3.5-turbo"
)
print(trimmed)
Alternatively, you can assign the token limit manually:
gpt_trim.trim(
messages,
max_tokens=100
)
You can also add system messages with ease:
import gpt_trim
messages = [
..., # long, long content
{
"role": "user",
"content": "It's about drive, it's about power"
}
]
trimmed = gpt_trim.advanced_trim(
messages,
system_messages=[
{
"role": "system",
"content": "You'll act like the celebrity: The Rock."
}
],
model="gpt-3.5-turbo",
)
print(trimmed)
The catch? It's slower. With great power comes great... patience.
You can compare this project to KillianLucas/tokentrim like so:
import time
import gpt_trim
import tiktoken
import tokentrim
pattern = "d!3h.l7$fj" # 10 tokens
messages = [
{
"role": "user",
"content": pattern * 5000 # 50000 tokens
}
]
# cache first
enc = tiktoken.get_encoding("cl100k_base")
gpt_trim.num_tokens_from_messages(
messages,
enc
)
def test(provider):
print("Testing", provider.__name__)
s = time.time()
result = provider.trim(
messages,
model="gpt-3.5-turbo",
)
print(f"took {(time.time() - s):.4f}s\n")
# Swap the following for every test and see tokentrim
# struggles when dealing with longer context.
test(gpt_trim)
test(tokentrim)
Right. I was bored.
gpt-trim's People
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. ๐๐๐
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google โค๏ธ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.