diff --git a/.github/workflows/docker.yaml b/.github/workflows/docker.yaml
index 2a8d2d0..87e9861 100644
--- a/.github/workflows/docker.yaml
+++ b/.github/workflows/docker.yaml
@@ -15,7 +15,7 @@ on:
 permissions: {}
 
 env:
-  VERSION: 0.1.6
+  VERSION: 0.1.7
   IMAGE_NAME: cpg-flow-seqr-loader
   DOCKER_DEV: australia-southeast1-docker.pkg.dev/cpg-common/images-dev
   DOCKER_MAIN: australia-southeast1-docker.pkg.dev/cpg-common/images
diff --git a/README.md b/README.md
index 91ecc94..604739d 100644
--- a/README.md
+++ b/README.md
@@ -55,7 +55,7 @@ CPG-Flow workflows are operated entirely by defining input Cohorts (see [here](h
 ```bash
 analysis-runner \
     --skip-repo-checkout \
-    --image australia-southeast1-docker.pkg.dev/cpg-common/images/cpg-flow-seqr-loader:0.1.6 \
+    --image australia-southeast1-docker.pkg.dev/cpg-common/images/cpg-flow-seqr-loader:0.1.7 \
     --config src/cpg_seqr_loader/config_template.toml \
     --config cohorts.toml \  # containing the inputs_cohorts and sequencing_type
     --dataset seqr \
@@ -70,7 +70,7 @@ analysis-runner \
 ```bash
 analysis-runner \
     --skip-repo-checkout \
-    --image australia-southeast1-docker.pkg.dev/cpg-common/images/cpg-flow-seqr-loader:0.1.6 \
+    --image australia-southeast1-docker.pkg.dev/cpg-common/images/cpg-flow-seqr-loader:0.1.7 \
     --config src/cpg_seqr_loader/config_template.toml \
     --config cohorts.toml \  # containing the inputs_cohorts and sequencing_type
     --dataset seqr \
diff --git a/pyproject.toml b/pyproject.toml
index c9d7a5d..faf2fbd 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -8,7 +8,7 @@ description='Seqr-Loader (gVCF-combiner) implemented in CPG-Flow'
 readme = "README.md"
 # currently cpg-flow is pinned to this version
 requires-python = ">=3.10,<3.11"
-version="0.1.6"
+version="0.1.7"
 license={"file" = "LICENSE"}
 classifiers=[
     'Environment :: Console',
@@ -122,7 +122,7 @@ hail = ["hail"]
 "src/cpg_seqr_loader/scripts/annotate_cohort.py" = ["E501"]
 
 [tool.bumpversion]
-current_version = "0.1.6"
+current_version = "0.1.7"
 parse = "(?P<major>\\d+)\\.(?P<minor>\\d+)\\.(?P<patch>\\d+)"
 serialize = ["{major}.{minor}.{patch}"]
 commit = true
diff --git a/src/cpg_seqr_loader/jobs/CombineGvcfsIntoVds.py b/src/cpg_seqr_loader/jobs/CombineGvcfsIntoVds.py
index 4a3767b..34d877b 100644
--- a/src/cpg_seqr_loader/jobs/CombineGvcfsIntoVds.py
+++ b/src/cpg_seqr_loader/jobs/CombineGvcfsIntoVds.py
@@ -3,7 +3,7 @@
 import loguru
 from cpg_flow import targets
 from cpg_flow import utils as cpg_flow_utils
-from cpg_utils import Path, config, hail_batch
+from cpg_utils import Path, config, hail_batch, to_path
 
 from cpg_seqr_loader import utils
 
@@ -15,13 +15,15 @@ def create_combiner_jobs(
     multicohort: targets.MultiCohort,
     output_vds: Path,
     combiner_plan: Path,
-    temp_dir: Path,
+    temp_dir_string: str,
     job_attrs: dict[str, str],
 ) -> 'BashJob | None':
     vds_path: str | None = None
     sg_ids_in_vds: set[str] = set()
     sgs_to_remove: list[str] = []
 
+    temp_dir = to_path(temp_dir_string)
+
     # check for a VDS by ID - this is not the typical RD process
     if vds_id := config.config_retrieve(['workflow', 'use_specific_vds'], None):
         vds_result_or_none = utils.query_for_specific_vds(vds_id)
diff --git a/src/cpg_seqr_loader/stages.py b/src/cpg_seqr_loader/stages.py
index da9f9d1..e1fc789 100644
--- a/src/cpg_seqr_loader/stages.py
+++ b/src/cpg_seqr_loader/stages.py
@@ -35,7 +35,7 @@ class CombineGvcfsIntoVds(stage.MultiCohortStage):
     def expected_outputs(self, multicohort: targets.MultiCohort) -> dict[str, Path | str]:
         return {
             'vds': self.prefix / f'{multicohort.name}.vds',
-            'tmp': self.tmp_prefix / 'temp_dir',
+            'tmp': str(self.tmp_prefix / 'temp_dir'),
         }
 
     def queue_jobs(self, multicohort: targets.MultiCohort, inputs: stage.StageInput) -> stage.StageOutput:
@@ -45,7 +45,7 @@ def queue_jobs(self, multicohort: targets.MultiCohort, inputs: stage.StageInput)
             multicohort=multicohort,
             output_vds=outputs['vds'],
             combiner_plan=self.tmp_prefix / 'combiner_plan.json',
-            temp_dir=outputs['tmp'],
+            temp_dir_string=outputs['tmp'],
             job_attrs=self.get_job_attrs(multicohort),
         )
         return self.make_outputs(multicohort, data=outputs, jobs=job)