HelpingAI-Lite
Subscribe to my YouTube channel
GGUF version here
HelpingAI-Lite is a lite version of the HelpingAI model that can assist with coding tasks. It's trained on a diverse range of datasets and fine-tuned to provide accurate and helpful responses.
License
This model is licensed under MIT.
Datasets
The model was trained on the following datasets:
- cerebras/SlimPajama-627B
- bigcode/starcoderdata
- HuggingFaceH4/ultrachat_200k
- HuggingFaceH4/ultrafeedback_binarized
Language
The model supports English language.
Usage
CPU and GPU code
from transformers import pipeline
from accelerate import Accelerator
# Initialize the accelerator
accelerator = Accelerator()
# Initialize the pipeline
pipe = pipeline("text-generation", model="OEvortex/HelpingAI-Lite", device=accelerator.device)
# Define the messages
messages = [
{
"role": "system",
"content": "You are a chatbot who can help code!",
},
{
"role": "user",
"content": "Write me a function to calculate the first 10 digits of the fibonacci sequence in Python and print it out to the CLI.",
},
]
# Prepare the prompt
prompt = pipe.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
# Generate predictions
outputs = pipe(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
# Print the generated text
print(outputs[0]["generated_text"])
- Downloads last month
- 509
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.
Model tree for OEvortex/HelpingAI-Lite
Datasets used to train OEvortex/HelpingAI-Lite
Evaluation results
- Epochself-reported3.000
- Eval Logits/Chosenself-reported-2.707
- Eval Logits/Rejectedself-reported-2.657
- Eval Logps/Chosenself-reported-370.130
- Eval Logps/Rejectedself-reported-296.074
- Eval Lossself-reported0.514
- Eval Rewards/Accuraciesself-reported0.738
- Eval Rewards/Chosenself-reported-0.027
- Eval Rewards/Marginsself-reported1.009
- Eval Rewards/Rejectedself-reported-1.036