d0ks/perry-anderson-lora

Perry Anderson Style LoRA

A LoRA adapter fine-tuned on the writing style of Perry Anderson, the British historian and essayist.

Model Details

• Base Model: moonshotai/Kimi-K2-Thinking
• LoRA Rank: 32
• Training: 5 epochs on ~100k words of Perry Anderson's essays
• Fine-tuning Platform: Tinker API

Writing Style

Perry Anderson's prose is characterized by:

• Long, complex sentences with multiple subordinate clauses
• Extensive use of historical parallels across centuries
• Sophisticated vocabulary drawing from multiple languages
• Marxist-inflected political and cultural analysis
• Dense, allusive prose that rewards careful reading

Usage

from peft import PeftModel
from transformers import AutoModelForCausalLM, AutoTokenizer

base_model = AutoModelForCausalLM.from_pretrained(
    "moonshotai/Kimi-K2-Thinking",
    trust_remote_code=True
)
model = PeftModel.from_pretrained(base_model, "YOUR_USERNAME/perry-anderson-lora")
tokenizer = AutoTokenizer.from_pretrained(
    "moonshotai/Kimi-K2-Thinking",
    trust_remote_code=True
)

# Generate text
inputs = tokenizer("The trajectory of European socialism", return_tensors="pt")
outputs = model.generate(**inputs, max_new_tokens=200)
print(tokenizer.decode(outputs[0]))

Training Details

Fine-tuned using causal language modeling on excerpts from Perry Anderson's essays and books, including works from New Left Review and Verso publications.