fix(llm_cost_calc/google.py): fix google embedding cost calculation

Fixes https://github.com/BerriAI/litellm/issues/4630
This commit is contained in:
Krrish Dholakia 2024-07-11 11:52:18 -07:00
parent db7d417727
commit 3f965df68b
6 changed files with 133 additions and 18 deletions

View file

@ -15,10 +15,12 @@ from litellm.litellm_core_utils.llm_cost_calc.google import (
from litellm.litellm_core_utils.llm_cost_calc.google import (
cost_per_token as google_cost_per_token,
)
from litellm.litellm_core_utils.llm_cost_calc.google import (
cost_router as google_cost_router,
)
from litellm.litellm_core_utils.llm_cost_calc.utils import _generic_cost_per_character
from litellm.types.llms.openai import HttpxBinaryResponseContent
from litellm.types.router import SPECIAL_MODEL_INFO_PARAMS
from litellm.utils import (
CallTypes,
CostPerToken,
@ -160,22 +162,32 @@ def cost_per_token(
# see this https://learn.microsoft.com/en-us/azure/ai-services/openai/concepts/models
print_verbose(f"Looking up model={model} in model_cost_map")
if custom_llm_provider == "vertex_ai" and "claude" in model:
return google_cost_per_token(
model=model_without_prefix,
custom_llm_provider=custom_llm_provider,
prompt_tokens=prompt_tokens,
completion_tokens=completion_tokens,
)
if custom_llm_provider == "vertex_ai":
return google_cost_per_character(
cost_router = google_cost_router(
model=model_without_prefix,
custom_llm_provider=custom_llm_provider,
prompt_characters=prompt_characters,
completion_characters=completion_characters,
prompt_tokens=prompt_tokens,
completion_tokens=completion_tokens,
call_type=call_type,
)
if cost_router == "cost_per_character":
return google_cost_per_character(
model=model_without_prefix,
custom_llm_provider=custom_llm_provider,
prompt_characters=prompt_characters,
completion_characters=completion_characters,
prompt_tokens=prompt_tokens,
completion_tokens=completion_tokens,
)
elif cost_router == "cost_per_token":
return google_cost_per_token(
model=model_without_prefix,
custom_llm_provider=custom_llm_provider,
prompt_tokens=prompt_tokens,
completion_tokens=completion_tokens,
)
elif custom_llm_provider == "gemini":
return google_cost_per_token(
model=model_without_prefix,