CodeGemma-2B-Slerp-dora is a DPO fine-tuned of johnsnowlabs/CodeGemma-2B-Slerp on argilla/distilabel-intel-orca-dpo-pairs preference dataset using DoRA. The model has been trained for 1080 steps. All hyperparams are given below.

🏆 Evaluation results

Coming Soom

Usage

python

!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "johnsnowlabs/CodeGemma-2B-dora"
messages = [{"role": "user", "content": "Explain what is Machine learning."}]

tokenizer = AutoTokenizer.from_pretrained(model)
prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
pipeline = transformers.pipeline(
    "text-generation",
    model=model,
    torch_dtype=torch.float16,
    device_map="auto",
)

outputs = pipeline(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
print(outputs[0]["generated_text"])

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 2e-04
train_batch_size: 1
gradient_accumulation_steps: 8
optimizer: PagedAdamW with 32-bit precision
lr_scheduler_type: Cosine
lr_scheduler_warmup_steps: 100
training_steps: 1080

LoRA Config

lora_r: 16
lora_alpha: 32
lora_dropout: 0.05
peft_use_dora: true

Framework versions

Transformers 4.39.0.dev0
Peft 0.9.1.dev0
Datasets 2.18.0
torch 2.2.0
accelerate 0.27.2

RichardErkhov/johnsnowlabs_-_CodeGemma-2B-dora-8bits

Original model description:

CodeGemma-2B-Slerp-dora