Update project parameters in train_llm.py

train_llm.py

@@ -31,7 +31,7 @@ dataset['validation'].to_pandas().to_csv(validation_filename, columns=['text'],
 # Define project parameters
 username='ai-aerospace'
 project_name='./llms/'+'ams_data_train-100_'+str(uuid4())
-repo_name='ams_data_train-100_'+str(uuid4())
+repo_name='ams-data-train-100-'+str(uuid4())
 
 # model_name='TinyLlama/TinyLlama-1.1B-Chat-v0.1'
 model_name='mistralai/Mistral-7B-v0.1'
@@ -42,16 +42,23 @@ os.environ["model_name"] = model_name
 os.environ["repo_id"] = username+'/'+repo_name
 os.environ["train_data"] = train_data   
 os.environ["validation_data"] = validation_data
+os.environ["data_directory"] = data_directory
+
+print("project_name:", project_name)
+print("model_name:", model_name)
+print("repo_id:", username+'/'+repo_name)
+print("train_data:", train_data)
+print("validation_data:", validation_data)
+print("data_directory:", data_directory)
 
 # Set .venv and execute the autotrain script
 # To see all parameters: autotrain llm --help
 # !autotrain llm --train --project_name my-llm --model TinyLlama/TinyLlama-1.1B-Chat-v0.1 --data_path . --use-peft --use_int4 --learning_rate 2e-4 --train_batch_size 6 --num_train_epochs 3 --trainer sft
-# The training dataset to be used must be called training.csv and be located in the data_path folder.
 command="""
 autotrain llm --train \
     --project_name ${project_name} \
     --model ${model_name} \
-    --data_path ../fine_tune_data \
+    --data_path ${data_directory} \
     --train_split ${train_data} \
     --valid_split ${validation_data} \
     --use-peft \