rebase

llama_stack/templates/open-benchmark/open_benchmark.py

@@ -215,8 +215,8 @@ def get_distribution_template() -> DistributionTemplate:
             purpose=DatasetPurpose.eval_messages_answer,
             source=URIDataSource(
                 uri="huggingface://datasets/llamastack/docvqa?split=val",
-            )
+            ),
-        )
+        ),
     ]
 
     default_benchmarks = [
@@ -254,7 +254,7 @@ def get_distribution_template() -> DistributionTemplate:
             benchmark_id="meta-reference-docvqa",
             dataset_id="docvqa",
             scoring_functions=["basic::docvqa"],
-        )
+        ),
     ]
     return DistributionTemplate(
         name=name,

llama_stack/templates/open-benchmark/run.yaml

@@ -188,6 +188,18 @@ datasets:
     uri: huggingface://datasets/llamastack/bfcl_v3?split=train
   metadata: {}
   dataset_id: bfcl
+- purpose: eval/messages-answer
+  source:
+    type: uri
+    uri: huggingface://datasets/llamastack/IfEval?split=train
+  metadata: {}
+  dataset_id: ifeval
+- purpose: eval/messages-answer
+  source:
+    type: uri
+    uri: huggingface://datasets/llamastack/docvqa?split=val
+  metadata: {}
+  dataset_id: docvqa
 scoring_fns: []
 benchmarks:
 - dataset_id: simpleqa
@@ -215,6 +227,16 @@ benchmarks:
   - basic::bfcl
   metadata: {}
   benchmark_id: meta-reference-bfcl
+- dataset_id: ifeval
+  scoring_functions:
+  - basic::ifeval
+  metadata: {}
+  benchmark_id: meta-reference-ifeval
+- dataset_id: docvqa
+  scoring_functions:
+  - basic::docvqa
+  metadata: {}
+  benchmark_id: meta-reference-docvqa
 tool_groups:
 - toolgroup_id: builtin::websearch
   provider_id: tavily-search