HumanFlow-Llama3-8B is a fine-tuned Llama 3 model designed to transform robotic AI-generated writing into content that feels natural, human, readable, and authentic.
HumanFlow rewrites outputs to feel more organic and naturally written.
1from transformers import AutoTokenizer, AutoModelForCausalLM
2import torch
3
4model_id = "randhir302/HumanFlow"
5
6tokenizer = AutoTokenizer.from_pretrained(model_id)
7
8model = AutoModelForCausalLM.from_pretrained(
9 model_id,
10 torch_dtype=torch.float16,
11 device_map="auto"
12)
13
14prompt = """
15Rewrite this in a more human tone:
16
17Artificial intelligence is transforming industries worldwide.
18"""
19
20inputs = tokenizer(prompt, return_tensors="pt").to("cuda")
21
22outputs = model.generate(
23 **inputs,
24 max_new_tokens=220,
25 temperature=0.75,
26 top_p=0.90,
27 repetition_penalty=1.10
28)
29
30print(tokenizer.decode(outputs[0], skip_special_tokens=True))
31
1temperature = 0.75
2top_p = 0.90
3repetition_penalty = 1.10
4max_new_tokens = 700
5
6## Roadmap
7
8- [x] Public Launch
9- [x] Hugging Face Release
10- [x] Fine-Tuned Base Model
11- [ ] GGUF Quantized Release
12- [ ] HumanFlow Pro API
13- [ ] Browser Editor
14- [ ] Multilingual Version
15
16---
17
18## Community
19
20If HumanFlow helps you:
21
22⭐ Like the model
23⭐ Share outputs
24⭐ Benchmark it
25⭐ Build products with it
26
