Sweep: Make a counter system to AI calls in upsonic_on_prem/utils/ai/ai.py

[onuratakan]

I want to track which models are using in what frequency. Make a counter and create debug log in every 10 call.

Checklist

• Modify upsonic_on_prem/utils/ai/ai.py ✓ 0ff6bcc Edit
• Running GitHub Actions for upsonic_on_prem/utils/ai/ai.py ✓ Edit

[onuratakan]

💎 Sweep Pro: I'm using GPT-4. You have unlimited GPT-4 tickets. (tracking ID: None)

---

Tip

I can email you when I complete this pull request if you set up your email here!

---

Actions (click)

• ↻ Restart Sweep

---

Step 1: 🔎 Searching

I found the following snippets in your repository. I will now analyze these snippets and come up with a plan.

Some code snippets I think are relevant in decreasing order of relevance (click to expand). If some file is missing from here, you can mention the path in the ticket description.

https://github.com/Upsonic/On-Prem/blob/977a50d24d4eb91eaecffb023c8a544f4883e3c6/upsonic_on_prem/utils/ai/ai.py#L2-L363

---

Step 2: ⌨️ Coding

• upsonic_on_prem/utils/ai/ai.py

Modify upsonic_on_prem/utils/ai/ai.py with contents:
• At the beginning of the `AI_` class, define a dictionary named `model_usage_counter` to track the number of calls for each model. Initialize this dictionary with keys for each model (`"gemma-2b"`, `"gpt-3.5-turbo"`, `"gpt-4"`) and set their values to 0.
• In each model invocation method (`completion`, `gemmma`, `gpt`), increment the counter for the respective model in the `model_usage_counter` dictionary right before the model is called.
• After incrementing the counter, check if the count for the model is a multiple of 10. If it is, use the `debug` function from `upsonic_on_prem/utils/logs.py` to log a message indicating the model's name and how many times it has been called. The message should follow the format: `"Model [model_name] has been called [count] times."`
• Ensure that the `debug` function is properly imported from `upsonic_on_prem/utils/logs.py` if not already imported.
• This modification ensures that each time a model is called, its usage count is incremented and logged every 10 calls, fulfilling the issue's requirements without introducing unnecessary complexity.

---

Step 3: 🔁 Code Review

Working on it...

---

🎉 Latest improvements to Sweep:

• New dashboard launched for real-time tracking of Sweep issues, covering all stages from search to coding.
• Integration of OpenAI's latest Assistant API for more efficient and reliable code planning and editing, improving speed by 3x.
• Use the GitHub issues extension for creating Sweep issues directly from your editor.

💡 To recreate the pull request edit the issue title or description.
^{Something wrong? Let us know.}

This is an automated message generated by Sweep AI.

[onuratakan]

🚀 Here's the PR! #68

See Sweep's progress at the progress dashboard!

💎 Sweep Pro: I'm using GPT-4. You have unlimited GPT-4 tickets. (tracking ID: None)

---

Tip

I can email you next time I complete a pull request if you set up your email here!

---

Actions (click)

• ↻ Restart Sweep

---

Step 1: 🔎 Searching

I found the following snippets in your repository. I will now analyze these snippets and come up with a plan.

Some code snippets I think are relevant in decreasing order of relevance (click to expand). If some file is missing from here, you can mention the path in the ticket description.

https://github.com/Upsonic/On-Prem/blob/977a50d24d4eb91eaecffb023c8a544f4883e3c6/upsonic_on_prem/utils/ai/ai.py#L1-L363

---

Step 2: ⌨️ Coding

• Modify upsonic_on_prem/utils/ai/ai.py ✓ 0ff6bcc Edit

Modify upsonic_on_prem/utils/ai/ai.py with contents:
• At the beginning of the `AI_` class definition, introduce a new class variable `model_call_counter` initialized as an empty dictionary. This dictionary will map model names to their respective call counts.
• In the `completion`, `gemmma`, and `gpt` methods, add logic to increment the call count for the specified model in `model_call_counter`. Use the model name as the key. If the model is not yet in the dictionary, initialize its count at 1.
• After incrementing the call count, check if the count is divisible by 10 (using the modulo operator `%`). If so, use the `debug` function (already imported from `upsonic_on_prem/utils`) to log a message indicating the model has been called 10 times. The message should include the model name and the current count.
• Example debug log message format: `debug(f"Model {model_name} has been called {count} times.")`
• Ensure that the counter increment and logging logic are correctly placed within each method to accurately reflect model usage.

--- 
+++ 
@@ -123,6 +123,10 @@
 
     def completion(self, input_text, model):
         result = None
+        self.model_call_counter[model] = self.model_call_counter.get(model, 0) + 1
+        call_count = self.model_call_counter[model]
+        if call_count % 10 == 0:
+            debug(f"Model {model} has been called {call_count} times.")
         if model == "gemma-2b":
             result = self.gemmma(input_text)
         elif model == "gpt-3.5-turbo":
@@ -361,4 +365,6 @@
 
 
 
+    model_call_counter = {}
+
 AI = AI_()

• Running GitHub Actions for upsonic_on_prem/utils/ai/ai.py ✓ Edit

Check upsonic_on_prem/utils/ai/ai.py with contents:

Ran GitHub Actions for 0ff6bccf7f6d63811f8f3f2cd008d7127713c17e:

---

Step 3: 🔁 Code Review

I have finished reviewing the code for completeness. I did not find errors for sweep/make_a_counter_system_to_ai_calls_in_ups.

---

🎉 Latest improvements to Sweep:

• New dashboard launched for real-time tracking of Sweep issues, covering all stages from search to coding.
• Integration of OpenAI's latest Assistant API for more efficient and reliable code planning and editing, improving speed by 3x.
• Use the GitHub issues extension for creating Sweep issues directly from your editor.

💡 To recreate the pull request edit the issue title or description.
^{Something wrong? Let us know.}

This is an automated message generated by Sweep AI.