GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
wildfellhall
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Run this model inference with full control and performance in your environment.
Model Details
Model Provider
wildfellhall
Model Tree
Input Modalities
Output Modalities
Supported Functionality
A small model (around 50 million parameters) trained on a dataset of poetry in the public domain.
Github Repository: https://github.com/wildfellhall/Poetry50M
Prompts work best in this format:
text
prompt = ( f"<|startofpoem|>\n" f"<|title|> {title}\n" f"<|author|> {author}\n" f"<|body|> \n")With temperature 0.4:
<|startofpoem|>
<|title|> Bereft
<|author|> Robert Frost
<|body|> . The world is gone, the earth is gone;
And there's a little while to-morrow.
The world has come, and every day
Is in its place of life and peace,
And all the world is done.
<|endofpoem|>
!pip install -U transformers from transformers import pipeline
pipe = pipeline("text-generation", model="wildfellhall/public-poetry-50m")
from transformers import AutoTokenizer, AutoModelForCausalLM
tokenizer = AutoTokenizer.from_pretrained("wildfellhall/public-poetry-50m") model = AutoModelForCausalLM.from_pretrained("wildfellhall/public-poetry-50m", device_map="auto")
from transformers import GPT2LMHeadModel, PreTrainedTokenizerFast import torch
saved_model = model.to("cuda" if torch.cuda.is_available() else "cpu")
title = input(" ") author = input(" ") prompt = ( f"<|startofpoem|>\n" f"<|title|> {title}\n" f"<|author|> {author}\n" f"<|body|> \n" )
inputs = tokenizer( prompt, return_tensors="pt").to(saved_model.device)
outputs = saved_model.generate( **inputs, max_new_tokens=150, temperature=0.8, top_p=0.9, repetition_penalty=1.15, do_sample=True, pad_token_id=tokenizer.pad_token_id, eos_token_id=tokenizer.eos_token_id )
print(tokenizer.decode(outputs[0], skip_special_tokens=False))