Downloads · 30 days
115
21% of all-time downloads
mesolitica/gemma-2B-16k-instructions
gemma-2B-16k-instructions is a text generation model from mesolitica. Use it when you need the model to write or continue text. It is set up for transformers.
README at https://github.com/mesolitica/malaya/tree/5.1/session/gemmainstructions-2b-16384-context-length
Downloads · 30 days
115
21% of all-time downloads
All-time downloads
536
Public
Parameters
2.5B
25.1 GB on disk
Likes
1
Public
Click a slice to open those files.
.safetensors5 GB · 100%
From the Hugging Face model README
README at https://github.com/mesolitica/malaya/tree/5.1/session/gemma#instructions-2b-16384-context-length
We use exact Gemma Instruct chat template.
WandB, https://wandb.ai/huseinzol05/gemma-2B-8192-fpf-instructions-16k?workspace=user-huseinzol05
Dataset gathered at https://huggingface.co/collections/mesolitica/malaysian-synthetic-dataset-656c2673fe7fe0b1e9e25fe2
Notebook to prepare dataset at https://github.com/mesolitica/malaysian-dataset/blob/master/llm-instruction/combine-malay-no-alignment-multitasks-v6.ipynb
This model is a quick demonstration that the base model can be easily fine-tuned to achieve some performance. It does have minimal moderation mechanisms.
from transformers import AutoTokenizer, AutoModelForCausalLM, BitsAndBytesConfig
import torch
import json
TORCH_DTYPE = 'bfloat16'
nf4_config = BitsAndBytesConfig(
load_in_4bit=True,
bnb_4bit_quant_type='nf4',
bnb_4bit_use_double_quant=True,
bnb_4bit_compute_dtype=getattr(torch, TORCH_DTYPE)
)
tokenizer = AutoTokenizer.from_pretrained('mesolitica/gemma-2B-16k-instructions')
model = AutoModelForCausalLM.from_pretrained(
'mesolitica/gemma-2B-16k-instructions',
use_flash_attention_2 = True,
quantization_config = nf4_config
)
messages = [
{'role': 'user', 'content': 'kwsp tu apa'}
]
prompt = tokenizer.apply_chat_template(messages, tokenize = False)
inputs = tokenizer([prompt], return_tensors='pt', add_special_tokens=False).to('cuda')
generate_kwargs = dict(
inputs,
max_new_tokens=1024,
top_p=0.95,
top_k=50,
temperature=0.9,
do_sample=True,
num_beams=1,
)
r = model.generate(**generate_kwargs)
tokenizer.decode(r[0])
<s> [INST] kwsp tu apa [/INST]KWSP bermaksud Kumpulan Wang Simpanan Pekerja. Ia adalah sebuah institusi simpanan persaraan yang ditubuhkan oleh Kementerian Kewangan Malaysia untuk tujuan mengumpul simpanan ahli untuk dibayar pada umur persaraan, penuh atau penuh persaraan penuh. KWSP ditubuhkan pada tahun 1951 dan mula beroperasi pada tahun 1952. KWSP adalah salah satu institusi simpanan persaraan terbesar di dunia, dengan pangkalan ahli sekitar 14 juta ahli.</s>