Gorilla BFCL patch update to automatize root cause detection and divergence in output between servings#4295
Open
dkalinowski wants to merge 1 commit into
Open
Gorilla BFCL patch update to automatize root cause detection and divergence in output between servings#4295dkalinowski wants to merge 1 commit into
dkalinowski wants to merge 1 commit into
Conversation
Contributor
There was a problem hiding this comment.
Pull request overview
This PR updates the Gorilla/BFCL patch used by the demos/continuous_batching/accuracy workflow to help automate root-cause detection of output divergence between different serving modes by optionally capturing server-side verbose response details into BFCL result metadata.
Changes:
- Adds a new
--include-verbose-logCLI flag and threads it through BFCL generation/inference flow. - Captures selected fields from a model server’s
__verboseresponse payload into BFCL result metadata for both multi-turn and single-turn inference. - Updates BFCL model registry entries for OVMS-backed evaluation (
ovms-model,ovms-model-stream).
| + include_verbose_log: bool = typer.Option( | ||
| + False, | ||
| + "--include-verbose-log", | ||
| + help="Include the __verbose field from model server responses (e.g. OVMS) in the result output; useful for debugging generation settings, prompts, and timings.", |
Comment on lines
+319
to
+322
| + if include_verbose_log and hasattr(api_response, "model_extra") and api_response.model_extra: | ||
| + verbose_data = api_response.model_extra.get("__verbose") | ||
| + if verbose_data: | ||
| + metadata["__verbose"] = {k: verbose_data[k] for k in ("prompt", "content") if k in verbose_data} |
Comment on lines
+337
to
+340
| + if include_verbose_log and hasattr(api_response, "model_extra") and api_response.model_extra: | ||
| + verbose_data = api_response.model_extra.get("__verbose") | ||
| + if verbose_data: | ||
| + metadata["__verbose"] = {k: verbose_data[k] for k in ("prompt", "content") if k in verbose_data} |
Comment on lines
+229
to
+233
| + if include_verbose_log and hasattr(api_response, "model_extra") and api_response.model_extra: | ||
| + verbose_data = api_response.model_extra.get("__verbose") | ||
| + if verbose_data: | ||
| + current_turn_verbose_log.append({k: verbose_data[k] for k in ("prompt", "content") if k in verbose_data}) | ||
| + |
Comment on lines
+283
to
+287
| + if include_verbose_log and hasattr(api_response, "model_extra") and api_response.model_extra: | ||
| + verbose_data = api_response.model_extra.get("__verbose") | ||
| + if verbose_data: | ||
| + current_turn_verbose_log.append({k: verbose_data[k] for k in ("prompt", "content") if k in verbose_data}) | ||
| + |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
No description provided.