Automatic dcft datacuration framework upload for mix_test

config/EvolInstruct/tests/evol_instruct_gpt-4o-mini_test.yaml

@@ -0,0 +1,57 @@
+operators:
+  - id: load_alpaca
+    config:
+      type: hf_source
+      dataset: yahma/alpaca-cleaned
+      split: train
+      columns: 
+        - instruction
+        - output
+      num_truncate: 100
+
+  - id: instruction_generation
+    config:
+      type: function
+      sharded: true
+      function: data_strategies.EvolInstruct.utils.instruction_generation
+      function_config:
+        input_column: instruction
+        output_column: evolved_instruction
+    input_ids:
+      - load_alpaca
+   
+  - id: annotate
+    input_ids:
+      - instruction_generation
+    config:
+      type: completions
+      map: chat
+      map_config:
+        user_message_column: evolved_instruction
+        output_column: completion
+      model: gpt-4o-mini
+      batch: False
+
+  - id: select_columns
+    config:
+      type: function
+      sharded: true
+      function: data_strategies.commons.select_columns
+      function_config:
+        columns:
+          - evolved_instruction
+          - completion
+    input_ids:
+      - annotate
+  
+  - id: to_sharegpt
+    config:
+      type: function
+      sharded: true
+      function: data_strategies.commons.convert_instruction_response_to_sharegpt
+      function_config:
+        input_instruction_column: evolved_instruction
+        input_response_column: completion
+        output_sharegpt_column: conversations
+    input_ids:
+      - select_columns