Upload test_iep_model.py with huggingface_hub

test_iep_model.py

@@ -0,0 +1,65 @@
+# /// script
+# requires-python = ">=3.10"
+# dependencies = [
+#     "transformers>=4.45.0",
+#     "peft>=0.7.0",
+#     "accelerate>=0.24.0",
+#     "torch",
+# ]
+# ///
+
+from peft import PeftModel
+from transformers import AutoModelForCausalLM, AutoTokenizer
+
+print("Loading base model and LoRA adapter...")
+base_model = AutoModelForCausalLM.from_pretrained(
+    "Qwen/Qwen3-4B",
+    torch_dtype="auto",
+    device_map="auto",
+)
+model = PeftModel.from_pretrained(base_model, "jeremierostan/iep-udl-qwen3-4b")
+tokenizer = AutoTokenizer.from_pretrained("jeremierostan/iep-udl-qwen3-4b")
+
+system_prompt = (
+    "You are an IEP specialist at ISP (International School of Panama). "
+    "You help teachers understand student accommodations, design inclusive instruction "
+    "using Universal Design for Learning (UDL) principles, and align teaching practices "
+    "with the ISP Way — the school's teaching and learning philosophy. The ISP Way "
+    "emphasizes learning as a journey of authentic discovery that is meaningful, exciting, "
+    "social, and iterative, with UDL as the framework for equity to remove barriers and "
+    "maintain high expectations for all learners."
+)
+
+test_questions = [
+    "What accommodations does student 2025-0347 receive for assessments?",
+    "How can I use UDL to support a student with dyslexia during a science lesson?",
+    "Which students are allowed breaks outside the classroom, and why?",
+    "I have a student with anxiety in my class. How can I design group activities that benefit all learners?",
+]
+
+for i, question in enumerate(test_questions, 1):
+    print(f"\n{'='*80}")
+    print(f"QUESTION {i}: {question}")
+    print(f"{'='*80}")
+
+    messages = [
+        {"role": "system", "content": system_prompt},
+        {"role": "user", "content": question},
+    ]
+
+    text = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True, enable_thinking=False)
+    inputs = tokenizer(text, return_tensors="pt").to(model.device)
+
+    output = model.generate(
+        **inputs,
+        max_new_tokens=512,
+        temperature=0.7,
+        top_p=0.9,
+        do_sample=True,
+    )
+
+    response = tokenizer.decode(output[0][inputs["input_ids"].shape[1]:], skip_special_tokens=True)
+    print(f"\nRESPONSE:\n{response}")
+
+print(f"\n{'='*80}")
+print("Testing complete!")