QLoRA: Efficient Finetuning of Quantized LLMs
Paper • 2305.14314 • Published • 61
# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("acrastt/Marx-3B-V3", trust_remote_code=True, dtype="auto")Quick Links
This is StableLM 3B 4E1T(Licensed under CC BY-SA 4.0.) finetuned on EverythingLM Data V3(ShareGPT Format) for 2 epochs using QLoRA.
Prompt template:
### HUMAN:
{prompt}
### RESPONSE:
Note that this model have the EOS token of <|endoftext|> instead of <\s>.
GGUF quantizations available here.
GPTQ quantizations available here.
- Downloads last month
- 34
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="acrastt/Marx-3B-V3", trust_remote_code=True)