fix: minor

.gitignore

@@ -47,6 +47,12 @@ owl/results
 logs/
 log/
 
+# Local files
+**/local_*
+**/local-*
+**/*_local*
+**/*-local*
+
 # Coverage reports
 htmlcov/
 .tox/

owl/scripts/prepare_sft.py

@@ -1,10 +1,10 @@
-from datasets import Dataset
+from datasets import Dataset, DatasetDict
 from tqdm import tqdm
 
 import json
+import random
 import argparse
 import pandas as pd
-import datasets
 
 
 def construct_system_message(task_prompt):
@@ -43,7 +43,6 @@ Please note that our overall task may be very complicated. Here are some tips th
     """
     
 
-
 def process_and_push(input_file, repo_id):
     r"""Process dataset and push to Hugging Face Hub."""
     
@@ -90,14 +89,20 @@ def process_and_push(input_file, repo_id):
     df["messages"] = df.apply(
         lambda row: build_messages(row["history"], row["question"]), axis=1)
     df["messages_camel"] = df["messages"]
+    
+    test_df = df.sample(n=min(5, len(df)), random_state=42)
+    train_df = df.drop(test_df.index)
+    
+    hf_dataset = DatasetDict({
+        "train": Dataset.from_pandas(train_df), 
+        "test": Dataset.from_pandas(test_df),
+    })
 
-    hf_dataset = Dataset.from_pandas(df)
     hf_dataset.push_to_hub(repo_id)
     print(f"Dataset pushed to: https://huggingface.co/datasets/{repo_id}")
 
 
 def main():
-    """Parse arguments and execute the processing."""
     parser = argparse.ArgumentParser(description="Process and push dataset.")
     parser.add_argument("--input_file", "-i", type=str, required=True,
                         help="Path to the input JSON file")
@@ -105,6 +110,7 @@ def main():
                         help="Hugging Face Hub repository ID")
     args = parser.parse_args()
 
+    random.seed(42)
     process_and_push(args.input_file, args.repo_id)
 
 

owl/scripts/serve.sh

@@ -5,6 +5,7 @@
 export VLLM_ALLOW_LONG_MAX_MODEL_LEN=1 
 
 model=${1:-Qwen/Qwen2.5-VL-7B-Instruct}
+# model=${1:-Qwen/Qwen2.5-7B-Instruct}
 max_model_len=${2:-32768} 
 gpu_memory_utilization=${3:-0.9}
 devices=${4:-"0,1,2,3"}

owl/scripts/sft.py

@@ -0,0 +1,96 @@
+"""
+Supervised Fine-Tuning (SFT) with owl agents' trajectories.
+"""
+
+import argparse
+
+from datasets import load_dataset
+from transformers import AutoConfig, AutoModelForCausalLM, AutoTokenizer
+from transformers.models.auto.modeling_auto import MODEL_FOR_IMAGE_TEXT_TO_TEXT_MAPPING_NAMES
+
+from trl import (
+    ModelConfig,
+    ScriptArguments,
+    SFTConfig,
+    SFTTrainer,
+    TrlParser,
+    get_kbit_device_map,
+    get_peft_config,
+    get_quantization_config,
+)
+
+
+def main(script_args, training_args, model_args):
+    ################
+    # Model init kwargs & Tokenizer
+    ################
+    quantization_config = get_quantization_config(model_args)
+    model_kwargs = dict(
+        revision=model_args.model_revision,
+        trust_remote_code=model_args.trust_remote_code,
+        attn_implementation=model_args.attn_implementation,
+        torch_dtype=model_args.torch_dtype,
+        use_cache=False if training_args.gradient_checkpointing else True,
+        device_map=get_kbit_device_map() if quantization_config is not None else None,
+        quantization_config=quantization_config,
+    )
+
+    # Create model
+    config = AutoConfig.from_pretrained(model_args.model_name_or_path)
+    model = AutoModelForCausalLM.from_pretrained(
+        model_args.model_name_or_path, 
+        **model_kwargs
+    )
+
+    # Create tokenizer
+    tokenizer = AutoTokenizer.from_pretrained(
+        model_args.model_name_or_path, trust_remote_code=model_args.trust_remote_code, use_fast=True
+    )
+    if tokenizer.pad_token is None:
+        tokenizer.pad_token = tokenizer.eos_token
+
+    ################
+    # Dataset
+    ################
+    dataset = load_dataset(
+        script_args.dataset_name, 
+        name=script_args.dataset_config
+    )
+
+    ################
+    # Training
+    ################
+    trainer = SFTTrainer(
+        model=model,
+        args=training_args,
+        train_dataset=dataset[script_args.dataset_train_split],
+        eval_dataset=dataset[script_args.dataset_test_split] if training_args.eval_strategy != "no" else None,
+        processing_class=tokenizer,
+        peft_config=get_peft_config(model_args),
+    )
+
+    trainer.train()
+
+    # Save and push to hub
+    trainer.save_model(training_args.output_dir)
+    if training_args.push_to_hub:
+        trainer.push_to_hub(dataset_name=script_args.dataset_name)
+
+
+def make_parser(subparsers: argparse._SubParsersAction = None):
+    dataclass_types = (ScriptArguments, SFTConfig, ModelConfig)
+    if subparsers is not None:
+        parser = subparsers.add_parser(
+            "sft", 
+            help="Run the SFT training script", 
+            dataclass_types=dataclass_types
+        )
+    else:
+        parser = TrlParser(dataclass_types)
+    return parser
+
+
+if __name__ == "__main__":
+    parser = make_parser()
+    script_args, training_args, model_args = parser.parse_args_and_config()
+    main(script_args, training_args, model_args)