Quantization made by Richard Erkhov.

Faro-Yi-34B-DPO - GGUF

Model creator: https://huggingface.co/wenbopan/
Original model: https://huggingface.co/wenbopan/Faro-Yi-34B-DPO/

Name	Quant method	Size
Faro-Yi-34B-DPO.Q2_K.gguf	Q2_K	11.94GB
Faro-Yi-34B-DPO.IQ3_XS.gguf	IQ3_XS	13.26GB
Faro-Yi-34B-DPO.IQ3_S.gguf	IQ3_S	13.99GB
Faro-Yi-34B-DPO.Q3_K_S.gguf	Q3_K_S	13.93GB
Faro-Yi-34B-DPO.IQ3_M.gguf	IQ3_M	14.5GB
Faro-Yi-34B-DPO.Q3_K.gguf	Q3_K	15.51GB
Faro-Yi-34B-DPO.Q3_K_M.gguf	Q3_K_M	15.51GB
Faro-Yi-34B-DPO.Q3_K_L.gguf	Q3_K_L	16.89GB
Faro-Yi-34B-DPO.IQ4_XS.gguf	IQ4_XS	17.36GB
Faro-Yi-34B-DPO.Q4_0.gguf	Q4_0	18.13GB
Faro-Yi-34B-DPO.IQ4_NL.gguf	IQ4_NL	18.3GB
Faro-Yi-34B-DPO.Q4_K_S.gguf	Q4_K_S	18.25GB
Faro-Yi-34B-DPO.Q4_K.gguf	Q4_K	19.24GB
Faro-Yi-34B-DPO.Q4_K_M.gguf	Q4_K_M	19.24GB
Faro-Yi-34B-DPO.Q4_1.gguf	Q4_1	20.1GB
Faro-Yi-34B-DPO.Q5_0.gguf	Q5_0	22.08GB
Faro-Yi-34B-DPO.Q5_K_S.gguf	Q5_K_S	22.08GB
Faro-Yi-34B-DPO.Q5_K.gguf	Q5_K	22.65GB
Faro-Yi-34B-DPO.Q5_K_M.gguf	Q5_K_M	22.65GB
Faro-Yi-34B-DPO.Q5_1.gguf	Q5_1	24.05GB
Faro-Yi-34B-DPO.Q6_K.gguf	Q6_K	26.28GB
Faro-Yi-34B-DPO.Q8_0.gguf	Q8_0	34.03GB

Original model description:

language: - en - zh license: mit datasets: - wenbopan/Chinese-dpo-pairs - Intel/orca_dpo_pairs - argilla/ultrafeedback-binarized-preferences-cleaned - jondurbin/truthy-dpo-v0.1 pipeline_tag: text-generation

Faro-Yi-9B-DPO

This is the DPO version of wenbopan/Faro-Yi-34B. Compared to Faro-Yi-34B and Yi-34B-200K, the DPO model excels at many tasks, surpassing the original Yi-34B-200K by a large margin.

How to Use

Faro-Yi-34B-DPO uses the chatml template and performs well in both short and long contexts.

import io
import requests
from PyPDF2 import PdfReader
from vllm import LLM, SamplingParams

llm = LLM(model="wenbopan/Faro-Yi-34B-DPO", kv_cache_dtype="fp8_e5m2", max_model_len=100000)

pdf_data = io.BytesIO(requests.get("https://arxiv.org/pdf/2303.08774.pdf").content)
document = "".join(page.extract_text() for page in PdfReader(pdf_data).pages) # 100 pages

question = f"{document}\n\nAccording to the paper, what is the parameter count of GPT-4?"
messages = [ {"role": "user", "content": question} ] # 83K tokens
prompt = llm.get_tokenizer().apply_chat_template(messages, add_generation_prompt=True, tokenize=False)
output = llm.generate(prompt, SamplingParams(temperature=0.8, max_tokens=500))
print(output[0].outputs[0].text)
# Yi-9B-200K:      175B. GPT-4 has 175B \nparameters. How many models were combined to create GPT-4? Answer: 6. ...
# Faro-Yi-9B: GPT-4 does not have a publicly disclosed parameter count due to the competitive landscape and safety implications of large-scale models like GPT-4. ...

Or With Transformers

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained('wenbopan/Faro-Yi-34B-DPO', device_map="cuda")
tokenizer = AutoTokenizer.from_pretrained('wenbopan/Faro-Yi-34B-DPO')
messages = [
    {"role": "system", "content": "You are a helpful assistant. Always answer with a short response."},
    {"role": "user", "content": "Tell me what is Pythagorean theorem like you are a pirate."}
]
input_ids = tokenizer.apply_chat_template(messages, tokenize=True, add_generation_prompt=True, return_tensors="pt").to(model.device)
generated_ids = model.generate(input_ids, max_new_tokens=512, temperature=0.5)
response = tokenizer.decode(generated_ids[0], skip_special_tokens=True) # Aye, matey! The Pythagorean theorem is a nautical rule that helps us find the length of the third side of a triangle. ...