README.md

metadata

tags:
  - autotrain
  - text-generation
widget:
  - text: 'I love AutoTrain because '
license: apache-2.0
datasets:
  - manjunathshiva/autotrain-data-GRADE3B-7B-02
language:
  - en

Model Trained Using AutoTrain

This model was trained using AutoTrain. For more information, please visit AutoTrain.

Usage

from transformers import AutoModelForCausalLM, AutoTokenizer
from peft import PeftModel
import torch
access_token = "<HF_TOKEN>"


tokenizer = AutoTokenizer.from_pretrained(
        "meta-llama/Llama-2-7b-chat-hf"
    )


base_model =  AutoModelForCausalLM.from_pretrained(
    'meta-llama/Llama-2-7b-chat-hf',
    token=access_token,
    trust_remote_code=True,
    #device_map="auto",    #Uncomment if you hava a good GPU Memory
    torch_dtype=torch.float16,
    offload_folder="offload/"
)
model = PeftModel.from_pretrained(
    base_model,
    'manjunathshiva/GRADE3B-7B-02-0',
    token=access_token,
    offload_folder="offload/"
    
).eval()

# Prompt content: "When is Maths Unit Test 2?"
messages = [
    {"role": "user", "content": "When is Maths Unit Test 2?"}
]

input_ids = tokenizer.apply_chat_template(conversation=messages, tokenize=True, add_generation_prompt=True, return_tensors='pt')
#output_ids = model.generate(input_ids.to('cuda'))  #Uncomment if you have CUDA and comment below line
output_ids = model.generate(input_ids=input_ids, temperature=0.01 )
response = tokenizer.decode(output_ids[0][input_ids.shape[1]:], skip_special_tokens=True)

# Model response: "<Outputs Date>"
print(response)