From dd1d0b561b982187f40486b1a4ea1ecbeb50a9d2 Mon Sep 17 00:00:00 2001 From: Jake Poznanski Date: Tue, 17 Jun 2025 17:06:45 +0000 Subject: [PATCH] More debug logs --- .github/workflows/main.yml | 2 ++ olmocr/pipeline.py | 14 ++++++++++++-- olmocr/version.py | 2 +- 3 files changed, 15 insertions(+), 3 deletions(-) diff --git a/.github/workflows/main.yml b/.github/workflows/main.yml index 3d833b1..7f010a1 100644 --- a/.github/workflows/main.yml +++ b/.github/workflows/main.yml @@ -276,6 +276,8 @@ jobs: with: token: ${{ secrets.BEAKER_TOKEN }} version: latest + - name: Debug Docker images + run: docker images - name: Push to Beaker env: diff --git a/olmocr/pipeline.py b/olmocr/pipeline.py index c2bfe64..dfd321c 100644 --- a/olmocr/pipeline.py +++ b/olmocr/pipeline.py @@ -1202,11 +1202,21 @@ async def main(): logger.info(f"Finished input tokens: {total_metrics.get('finished_input_tokens', 0):,}") logger.info(f"Finished output tokens: {total_metrics.get('finished_output_tokens', 0):,}") + logger.info(f"Completed pages: {total_metrics.get('completed_pages', 0):,}") + logger.info(f"Failed pages: {total_metrics.get('failed_pages', 0):,}") + logger.info( + f"Page Failure rate: {total_metrics.get('failed_pages', 0) / max(total_metrics.get('completed_pages', 0) + total_metrics.get('failed_pages', 0), 1) * 100:.2f}%" + ) + # Output rates if "server_input_tokens_per_sec" in rates: - logger.info(f"Input tokens/sec rate: {rates['server_input_tokens_per_sec']:.2f}") + logger.info(f"Server Input tokens/sec rate: {rates['server_input_tokens_per_sec']:.2f}") if "server_output_tokens_per_sec" in rates: - logger.info(f"Output tokens/sec rate: {rates['server_output_tokens_per_sec']:.2f}") + logger.info(f"Server Output tokens/sec rate: {rates['server_output_tokens_per_sec']:.2f}") + if "finished_input_tokens_per_sec" in rates: + logger.info(f"Finished Input tokens/sec rate: {rates['finished_input_tokens_per_sec']:.2f}") + if "finished_output_tokens_per_sec" in rates: + logger.info(f"Finished Output tokens/sec rate: {rates['finished_output_tokens_per_sec']:.2f}") logger.info("=" * 80) logger.info("Work done") diff --git a/olmocr/version.py b/olmocr/version.py index 94fdcdf..f5036aa 100644 --- a/olmocr/version.py +++ b/olmocr/version.py @@ -2,7 +2,7 @@ _MAJOR = "0" _MINOR = "1" # On main and in a nightly release the patch should be one ahead of the last # released build. -_PATCH = "72" +_PATCH = "73" # This is mainly for nightly builds which have the suffix ".dev$DATE". See # https://semver.org/#is-v123-a-semantic-version for the semantics. _SUFFIX = ""