Views
No views yet
"batch_size": 32,
"block_size": 256,
"lr": 5e-4,
"num_hidden_layers": 6,
"num_attention_heads": 6,
"hidden_size": 288,
"dropout": 0.1,
"weight_decay": 0.01,
"epochs": 1,
"eval_interval": 200,
"eval_steps": 50,
"vocab_size": 50257,
"warmup_tokens": 10000,
"gradient_accumulation_steps": 8,1 !pip install --quiet transformers
2 from transformers import AutoModelForCausalLM, AutoTokenizer
3 from huggingface_hub import notebook_login, login
4 import os
5
6 #login to hf to check for llama access
7 hf_token = os.getenv('HF_TOKEN')
8 login(token=hf_token)
9
10 model = AutoModelForCausalLM.from_pretrained('AnirudhRajagopalan1201/tinyllama-37M')
11 tokenizer = AutoTokenizer.from_pretrained("meta-llama/Llama-2-7b-hf")
12 prompt = "Lily likes cats and dogs. She asked her mom for a dog and her mom said no, so instead she asked"
13 input_ids = tokenizer.encode(prompt, return_tensors="pt")
14 output = model.generate(input_ids, temperature=0.1, max_length = 100, do_sample=True)
15 output_text = tokenizer.decode(output[0], skip_special_tokens=True)
16 print(output_text)
17