make sure to register the base chatml template even if no system mess…

…age is provided (#1207)

.github/workflows/tests.yml

@@ -106,3 +106,7 @@ jobs:
       - name: GPU Unit Tests monkeypatched w docker image
         run: |
           docker run --privileged --gpus "all" --env WANDB_DISABLED=true --rm ${{ steps.metadata.outputs.tags }}-py${{ matrix.python_version }}-cu${{ matrix.cuda }}-${{ matrix.pytorch }} pytest /workspace/axolotl/tests/e2e/patched/
+      - name: Prune image from docker
+        if: github.ref != 'refs/heads/main'
+        run: |
+          docker rmi -f ${{ steps.metadata.outputs.tags }}-py${{ matrix.python_version }}-cu${{ matrix.cuda }}-${{ matrix.pytorch }}

src/axolotl/cli/preprocess.py

@@ -40,6 +40,8 @@ def do_cli(config: Path = Path("examples/"), **kwargs):
             f"ChatML set. Adding default system message: {parsed_cfg.default_system_message}"
         )
         register_chatml_template(parsed_cfg.default_system_message)
+    else:
+        register_chatml_template()
 
     if not parsed_cfg.dataset_prepared_path:
         msg = (

src/axolotl/cli/train.py

@@ -43,7 +43,10 @@ def do_train(cfg, cli_args) -> Tuple[PreTrainedModel, PreTrainedTokenizer]:
             f"ChatML set. Adding default system message: {cfg.default_system_message}"
         )
         register_chatml_template(cfg.default_system_message)
+    else:
+        register_chatml_template()
 
+    if cfg.rl:
         dataset_meta = load_rl_datasets(cfg=cfg, cli_args=cli_args)
     else:
         dataset_meta = load_datasets(cfg=cfg, cli_args=cli_args)

src/axolotl/utils/data.py

@@ -16,6 +16,7 @@
     load_from_disk,
 )
 from huggingface_hub import hf_hub_download
+from huggingface_hub.utils import HFValidationError
 from torch.utils.data import RandomSampler
 from transformers import PreTrainedTokenizerBase
 
@@ -213,7 +214,7 @@ def for_d_in_datasets(dataset_configs):
                     token=use_auth_token,
                 )
                 ds_from_hub = True
-            except (FileNotFoundError, ConnectionError):
+            except (FileNotFoundError, ConnectionError, HFValidationError):
                 pass
 
             ds_from_cloud = False