Views
No views yet
Research Paper ──► [T5-Base-Summarization] ──► Summary ──► [T5-Base-Story-Generation] ──► Story| Parameter | Value |
|---|---|
| Base model | harsharajkumar273/T5-Base-Summarization |
| Task | Story Generation |
| Max input length | 512 tokens |
| Max target length | 256 tokens |
| Learning rate | 1e-4 |
| Batch size | 4 |
| Gradient accumulation steps | 4 |
| Warmup steps | 500 |
| Weight decay | 0.01 |
| Fine-tuning method | LoRA (r=16, alpha=32, targets: q, v) |
1from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
2
3# Stage 1: Summarize the paper
4sum_tokenizer = AutoTokenizer.from_pretrained("harsharajkumar273/T5-Base-Summarization")
5sum_model = AutoModelForSeq2SeqLM.from_pretrained("harsharajkumar273/T5-Base-Summarization")
6
7paper_text = "Your research paper text here..."
8word_count = len(paper_text.split())
9sum_prompt = f"Summarize this part of the research paper to less than {word_count // 10} words:\n{paper_text}"
10sum_inputs = sum_tokenizer(sum_prompt, return_tensors="pt", max_length=1024, truncation=True)
11sum_outputs = sum_model.generate(**sum_inputs, max_length=128, num_beams=4)
12summary = sum_tokenizer.decode(sum_outputs[0], skip_special_tokens=True)
13
14# Stage 2: Generate a story from the summary
15story_tokenizer = AutoTokenizer.from_pretrained("harsharajkumar273/T5-Base-Story-Generation")
16story_model = AutoModelForSeq2SeqLM.from_pretrained("harsharajkumar273/T5-Base-Story-Generation")
17
18story_inputs = story_tokenizer(summary, return_tensors="pt", max_length=512, truncation=True)
19story_outputs = story_model.generate(**story_inputs, max_length=256, num_beams=4)
20story = story_tokenizer.decode(story_outputs[0], skip_special_tokens=True)
21print(story)