phoenix/litellm
1
0
Fork 0
forked from phoenix/litellm-mirror
Code Pull requests Activity

docs update

Browse source
This commit is contained in:
ishaan-jaff 2023-09-06 17:16:24 -07:00
parent 1ba6b6761b
commit bd77d5ac21
3 changed files with 17 additions and 13 deletions
Download patch file Download diff file Expand all files Collapse all files

20
docs/my-website/docs/token_usage.md
View file

@ -1,4 +1,4 @@
# Token Usage # Completion Token Usage & Cost
By default LiteLLM returns token usage in all completion requests ([See here](https://litellm.readthedocs.io/en/latest/output/)) By default LiteLLM returns token usage in all completion requests ([See here](https://litellm.readthedocs.io/en/latest/output/))
However, we also expose 3 public helper functions to calculate token usage across providers: However, we also expose 3 public helper functions to calculate token usage across providers:
@ -33,13 +33,19 @@ print(prompt_tokens_cost_usd_dollar, completion_tokens_cost_usd_dollar)
``` ```
3. `completion_cost` 3. `completion_cost`
Accepts a `litellm.completion()` response and return a `float` of cost for the `completion` call
```python ```python
from litellm import completion_cost from litellm import completion, completion_cost
prompt = "Hey, how's it going" response = completion(
completion = "Hi, I'm gpt - I am doing well" model="together_ai/togethercomputer/llama-2-70b-chat",
cost_of_query = completion_cost(model="gpt-3.5-turbo", prompt=prompt, completion=completion)) messages=messages,
request_timeout=200,
print(cost_of_query) )
# pass your response from completion to completion_cost
cost = completion_cost(completion_response=response)
formatted_string = f"${float(cost):.10f}"
print(formatted_string)
``` ```

4
litellm/tests/test_completion_cost.py
View file

@ -28,11 +28,11 @@ def test_completion_togetherai_cost():
) )
# Add any assertions here to check the response # Add any assertions here to check the response
print(response) print(response)
print("Completion Cost:") print("Completion Cost: for togethercomputer/llama-2-70b-chat")
cost = completion_cost(completion_response=response) cost = completion_cost(completion_response=response)
formatted_string = f"${float(cost):.10f}" formatted_string = f"${float(cost):.10f}"
print(formatted_string) print(formatted_string)
except Exception as e: except Exception as e:
pytest.fail(f"Error occurred: {e}") pytest.fail(f"Error occurred: {e}")
# test_completion_togetherai_cost() test_completion_togetherai_cost()

6
litellm/utils.py
View file

@ -578,9 +578,7 @@ def get_model_params_and_category(model_name):
if params_match != None: if params_match != None:
params_match = params_match.group(1) params_match = params_match.group(1)
params_match = params_match.replace("b", "") params_match = params_match.replace("b", "")
print(params_match)
params_billion = float(params_match) params_billion = float(params_match)
# Determine the category based on the number of parameters # Determine the category based on the number of parameters
if params_billion <= 3.0: if params_billion <= 3.0:
category = "together-ai-up-to-3b" category = "together-ai-up-to-3b"
@ -593,8 +591,8 @@ def get_model_params_and_category(model_name):
elif params_billion <= 70.0: elif params_billion <= 70.0:
category = "together-ai-40.1b-70b" category = "together-ai-40.1b-70b"
return category return category
return "Model name not recognized or category not found." return None
def token_counter(model, text): def token_counter(model, text):