Update src/display/about.py

src/display/about.py

@@ -40,7 +40,23 @@ LLM_BENCHMARKS_TEXT = f"""
 ## How it works
 
 ## Reproducibility
-To reproduce our results, here is the commands you can run:
+To reproduce my results, here is the commands you can run:
+
+I use LM-Evaluation-Harness-Turkish, a version of the LM Evaluation Harness adapted for Turkish datasets, to ensure our leaderboard results are both reliable and replicable. Please see https://github.com/malhajar17/lm-evaluation-harness_turkish for more information
+
+## How to Reproduce Results:
+
+### 1) Set Up the repo: Clone the "lm-evaluation-harness_turkish" @ https://github.com/malhajar17/lm-evaluation-harness_turkish and follow the installation instructions.
+### 2) Run Evaluations: To get the results as on the leaderboard (Some tests might show small variations), use the following command, adjusting for your model. For example, with the Trendyol model:
+```python
+lm_eval --model vllm --model_args pretrained=Trendyol/Trendyol-LLM-7b-chat-v1.0 --tasks truthfulqa_mc2_tr,truthfulqa_mc1_tr,mmlu_tr,winogrande_tr,gsm8k_tr,arc_challenge_tr,hellaswag_tr --output /workspace/Trendyol-LLM-7b-chat-v1.0
+```
+### 3) Report Results: I take the average of truthfulqa_mc1_tr and truthfulqa_mc2_tr scores and report it as truthfulqa. The results file generated is then uploaded to the OpenLLM Turkish Leaderboard.
+
+Notes:
+
+- I currently use "vllm" which might differ slightly as per the LM Evaluation Harness.
+- All the tests are using "acc" as metric, with a plan to migrate to "acc_norm" for "ARC" and "Hellaswag" soon. 
 
 """