Model Card for Model ID
Model Details
Model Description
Model Based in Amazon music instruments dataset Reccomendation by prompt task
T5-recs
T5 with prompt Tuning PEFT finetuned on Amazon purchased data
Model source: https://huggingface.co/Fidlobabovic/T5-recs
Usagefrom transformers import AutoModelForCausalLM, AutoTokenizer
from peft import get_peft_config, get_peft_model, LoraConfig, TaskType
model = AutoModelForSeq2SeqLM.from_pretrained("t5-base")
model = PeftModel.from_pretrained(model, "Fidlobabovic/T5-recs")
tokenizer = AutoTokenizer.from_pretrained("Fidlobabovic/T5-recs")
input_text = "Purchases: { Guitar, Stradivary notes, synthesizer} Candidates: {Violin; Thrombon; Refrigirator; Sex toy;} - RECCOMENDATION :"
inputs = tokenizer(input_text, return_tensors="pt")
outputs = model.generate(input_ids=inputs["input_ids"], max_new_tokens=5)
print("input user hitory: ", input_text) ## output rec: ['Guitar']
print(" output rec: ", tokenizer.batch_decode(outputs.detach().cpu().numpy(), skip_special_tokens=True))
- Downloads last month
- 59
Model tree for Fidlobabovic/T5-recs
Base model
google-t5/t5-base