Skip to content

OpenAI model cost based on model name fix #1946

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

Open
wants to merge 4 commits into
base: main
Choose a base branch
from
Open

OpenAI model cost based on model name fix #1946

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
734b0f5
add model name to the TokenUsage object
JonasElburgUVA Mar 4, 2025
72e7806
retain model when adding TokenUsage objects together
JonasElburgUVA Mar 4, 2025
df779fe
Add model ID's to other parsers and update tests to expect a model va…
JonasElburgUVA Mar 7, 2025
308b53c
Fallback empty string as model ID if the llm results do not come with…
JonasElburgUVA Mar 7, 2025
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
25 changes: 20 additions & 5 deletions src/ragas/cost.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -22,6 +22,7 @@ def __add__(self, y: "TokenUsage") -> "TokenUsage":
return TokenUsage(
input_tokens=self.input_tokens + y.input_tokens,
output_tokens=self.output_tokens + y.output_tokens,
model=self.model
)
else:
raise ValueError("Cannot add TokenUsage objects with different models")
Expand Down Expand Up @@ -67,8 +68,9 @@ def get_token_usage_for_openai(
return TokenUsage(input_tokens=0, output_tokens=0)
output_tokens = get_from_dict(llm_output, "token_usage.completion_tokens", 0)
input_tokens = get_from_dict(llm_output, "token_usage.prompt_tokens", 0)
model = get_from_dict(llm_output, "model_name", "")

return TokenUsage(input_tokens=input_tokens, output_tokens=output_tokens)
return TokenUsage(input_tokens=input_tokens, output_tokens=output_tokens, model=model)


def get_token_usage_for_anthropic(
Expand All @@ -92,10 +94,16 @@ def get_token_usage_for_anthropic(
"usage.output_tokens",
0,
),
model=get_from_dict(
g.message.response_metadata,
"model",
"")
)
)

return sum(token_usages, TokenUsage(input_tokens=0, output_tokens=0))
model = next(
(usage.model for usage in token_usages if usage.model), ""
)
return sum(token_usages, TokenUsage(input_tokens=0, output_tokens=0, model=model))
else:
return TokenUsage(input_tokens=0, output_tokens=0)

Expand All @@ -120,10 +128,17 @@ def get_token_usage_for_bedrock(
"usage.completion_tokens",
0,
),
model=get_from_dict(
g.message.response_metadata,
"model_id",
""
)
)
)

return sum(token_usages, TokenUsage(input_tokens=0, output_tokens=0))
model = next(
(usage.model for usage in token_usages if usage.model), ""
)
return sum(token_usages, TokenUsage(input_tokens=0, output_tokens=0, model=model))
return TokenUsage(input_tokens=0, output_tokens=0)


Expand Down
10 changes: 5 additions & 5 deletions tests/unit/test_cost.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -133,27 +133,27 @@ def test_token_usage_cost():
def test_parse_llm_results():
# openai
token_usage = get_token_usage_for_openai(openai_llm_result)
assert token_usage == TokenUsage(input_tokens=10, output_tokens=10)
assert token_usage == TokenUsage(input_tokens=10, output_tokens=10, model="gpt-4o")

# anthropic
token_usage = get_token_usage_for_anthropic(anthropic_llm_result)
assert token_usage == TokenUsage(input_tokens=9, output_tokens=12)
assert token_usage == TokenUsage(input_tokens=9, output_tokens=12, model="claude-3-opus-20240229")

# Bedrock LLaMa
token_usage = get_token_usage_for_bedrock(bedrock_llama_result)
assert token_usage == TokenUsage(input_tokens=10, output_tokens=10)
assert token_usage == TokenUsage(input_tokens=10, output_tokens=10, model="us.meta.llama3-1-70b-instruct-v1:0")

# Bedrock Claude
token_usage = get_token_usage_for_bedrock(bedrock_claude_result)
assert token_usage == TokenUsage(input_tokens=10, output_tokens=10)
assert token_usage == TokenUsage(input_tokens=10, output_tokens=10, model="us.anthropic.claude-3-5-sonnet-20240620-v1:0")


def test_cost_callback_handler():
cost_cb = CostCallbackHandler(token_usage_parser=get_token_usage_for_openai)
cost_cb.on_llm_end(openai_llm_result)

# cost
assert cost_cb.total_tokens() == TokenUsage(input_tokens=10, output_tokens=10)
assert cost_cb.total_tokens() == TokenUsage(input_tokens=10, output_tokens=10, model="gpt-4o")

assert cost_cb.total_cost(0.1) == 2.0
assert (
Expand Down
Loading