metadata
license: apache-2.0
base_model: unsloth/qwen2.5-1.5b-instruct-unsloth-bnb-4bit
library_name: peft
tags:
- agriculture
- india
- advisory
- qwen
- qwen2.5
- lora
- krishi-anuman-x
- sih25099
language:
- en
- hi
pipeline_tag: text-generation
KRISHI-ANUMAN-X Advisory LLM (LoRA)
LoRA adapters fine-tuned on ICAR/IPM farm advisories for Qwen2.5-1.5B-Instruct.
Load
from peft import AutoPeftModelForCausalLM
from transformers import AutoTokenizer
model = AutoPeftModelForCausalLM.from_pretrained("Veeda241/krishi-anuman-qwen-1.5b")
tokenizer = AutoTokenizer.from_pretrained("Veeda241/krishi-anuman-qwen-1.5b")
Or merge locally, then serve with vLLM/Ollama and set:
QWEN_BASE_URL=http://localhost:8000/v1
QWEN_MODEL=...
See project agentic_ai/qwen/finetune/README.md.