RichardErkhov/frankenmerger_-_delta-4B-super-gguf

Quantization made by Richard Erkhov.

delta-4B-super - GGUF

Model creator: https://huggingface.co/frankenmerger/
Original model: https://huggingface.co/frankenmerger/delta-4B-super/

Name	Quant method	Size
delta-4B-super.Q2_K.gguf	Q2_K	1.7GB
delta-4B-super.IQ3_XS.gguf	IQ3_XS	1.84GB
delta-4B-super.IQ3_S.gguf	IQ3_S	1.92GB
delta-4B-super.Q3_K_S.gguf	Q3_K_S	1.92GB
delta-4B-super.IQ3_M.gguf	IQ3_M	2.04GB
delta-4B-super.Q3_K.gguf	Q3_K	2.21GB
delta-4B-super.Q3_K_M.gguf	Q3_K_M	2.21GB
delta-4B-super.Q3_K_L.gguf	Q3_K_L	2.45GB
delta-4B-super.IQ4_XS.gguf	IQ4_XS	2.38GB
delta-4B-super.Q4_0.gguf	Q4_0	2.48GB
delta-4B-super.IQ4_NL.gguf	IQ4_NL	2.51GB
delta-4B-super.Q4_K_S.gguf	Q4_K_S	2.51GB
delta-4B-super.Q4_K.gguf	Q4_K	2.7GB
delta-4B-super.Q4_K_M.gguf	Q4_K_M	2.7GB
delta-4B-super.Q4_1.gguf	Q4_1	2.75GB
delta-4B-super.Q5_0.gguf	Q5_0	3.01GB
delta-4B-super.Q5_K_S.gguf	Q5_K_S	3.01GB
delta-4B-super.Q5_K.gguf	Q5_K	3.13GB
delta-4B-super.Q5_K_M.gguf	Q5_K_M	3.13GB
delta-4B-super.Q5_1.gguf	Q5_1	3.01GB
delta-4B-super.Q6_K.gguf	Q6_K	3.57GB
delta-4B-super.Q8_0.gguf	Q8_0	4.63GB

Original model description:

widget: - text: Hello, My name is Junpei Iori, who are you? example_title: Identity - text: Describe Aurora Borealis example_title: Capabilities - text: Create a fastapi endpoint to retrieve the weather given a zip code. example_title: Coding pipeline_tag: text-generation license: apache-2.0 language: - en tags: - conversational

delta-4b-super is a frankenmerge of Phi-2-super.

💻 Usage

!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "gmonsoon/Delta-4B-Base"
messages = [{"role": "user", "content": "What is a large language model?"}]

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"])