| #!/bin/bash |
| set -e |
| |
| cd "$(dirname "$0")/../.." |
|
|
| OUTPUT_DIR="data/eval_results_thinking" |
|
|
| echo "==========================================" |
| echo "RUNNING ALL EVALS — parallelized by model" |
| echo "==========================================" |
|
|
| |
| ( |
| python3 sec4_3_improvements/extended_reasoning/eval_script_thinking.py --model gpt-5.4 --output_dir $OUTPUT_DIR |
| python3 sec4_3_improvements/extended_reasoning/eval_transparent_thinking.py --model gpt-5.4 --output_dir $OUTPUT_DIR |
| ) & |
|
|
| |
| ( |
| python3 sec4_3_improvements/extended_reasoning/eval_script_thinking.py --model claude-opus-4-6 --output_dir $OUTPUT_DIR |
| python3 sec4_3_improvements/extended_reasoning/eval_transparent_thinking.py --model claude-opus-4-6 --output_dir $OUTPUT_DIR |
| ) & |
|
|
| wait |
|
|
| echo "" |
| echo "==========================================" |
| echo "JUDGING" |
| echo "==========================================" |
|
|
| python3 sec3_occludebench/judger_script.py --results_dir $OUTPUT_DIR --output_dir $OUTPUT_DIR --judge_models gpt-5.4 claude-opus-4-6 |
|
|
| echo "" |
| echo "==========================================" |
| echo "ALL DONE — results in $OUTPUT_DIR/" |
| echo "==========================================" |
