Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Getting context window sizes of models without prefixes #994

Merged
merged 2 commits into from
May 13, 2024
Merged

Getting context window sizes of models without prefixes #994

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
64ed824
fix: getting context window sizes of models without prefixes
elisalimli May 13, 2024
23e3ae9
feat: limit split counts to 1
elisalimli May 13, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
23 changes: 20 additions & 3 deletions libs/superagent/app/memory/buffer_memory.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,7 +7,23 @@
from app.memory.message import BaseMessage

DEFAULT_TOKEN_LIMIT_RATIO = 0.75
DEFAULT_TOKEN_LIMIT = 3000
DEFAULT_TOKEN_LIMIT = 3072


def get_context_window(model: str) -> int:
max_input_tokens = model_cost.get(model, {}).get("max_input_tokens")

# Some models don't have a provider prefix in their name
# But they point to the same model
# Example: claude-3-haiku-20240307 and anthropic/claude-3-haiku-20240307
if not max_input_tokens:
model_parts = model.split("/", 1)
if len(model_parts) > 1:
model_without_prefix = model_parts[1]
max_input_tokens = model_cost.get(model_without_prefix, {}).get(
"max_input_tokens", DEFAULT_TOKEN_LIMIT
)
return max_input_tokens


class BufferMemory(BaseMemory):
Expand All @@ -21,8 +37,9 @@ def __init__(
self.memory_store = memory_store
self.tokenizer_fn = tokenizer_fn
self.model = model
context_window = model_cost.get(self.model, {}).get("max_input_tokens")
self.context_window = max_tokens or context_window * DEFAULT_TOKEN_LIMIT_RATIO
self.context_window = (
max_tokens or get_context_window(model=model) * DEFAULT_TOKEN_LIMIT_RATIO
)

def add_message(self, message: BaseMessage) -> None:
self.memory_store.add_message(message)
Expand Down