metadata
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
model-index:
- name: Faro-Yi-34B-DPO
results:
- task:
type: text-generation
name: Text Generation
dataset:
name: ENEM Challenge (No Images)
type: eduagarcia/enem_challenge
split: train
args:
num_few_shot: 3
metrics:
- type: acc
value: 73.55
name: accuracy
source:
url: >-
https://huggingface.co/spaces/eduagarcia/open_pt_llm_leaderboard?query=wenbopan/Faro-Yi-34B-DPO
name: Open Portuguese LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: BLUEX (No Images)
type: eduagarcia-temp/BLUEX_without_images
split: train
args:
num_few_shot: 3
metrics:
- type: acc
value: 65.79
name: accuracy
source:
url: >-
https://huggingface.co/spaces/eduagarcia/open_pt_llm_leaderboard?query=wenbopan/Faro-Yi-34B-DPO
name: Open Portuguese LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: OAB Exams
type: eduagarcia/oab_exams
split: train
args:
num_few_shot: 3
metrics:
- type: acc
value: 55.85
name: accuracy
source:
url: >-
https://huggingface.co/spaces/eduagarcia/open_pt_llm_leaderboard?query=wenbopan/Faro-Yi-34B-DPO
name: Open Portuguese LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: Assin2 RTE
type: assin2
split: test
args:
num_few_shot: 15
metrics:
- type: f1_macro
value: 92.2
name: f1-macro
source:
url: >-
https://huggingface.co/spaces/eduagarcia/open_pt_llm_leaderboard?query=wenbopan/Faro-Yi-34B-DPO
name: Open Portuguese LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: Assin2 STS
type: eduagarcia/portuguese_benchmark
split: test
args:
num_few_shot: 15
metrics:
- type: pearson
value: 79.78
name: pearson
source:
url: >-
https://huggingface.co/spaces/eduagarcia/open_pt_llm_leaderboard?query=wenbopan/Faro-Yi-34B-DPO
name: Open Portuguese LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: FaQuAD NLI
type: ruanchaves/faquad-nli
split: test
args:
num_few_shot: 15
metrics:
- type: f1_macro
value: 71
name: f1-macro
source:
url: >-
https://huggingface.co/spaces/eduagarcia/open_pt_llm_leaderboard?query=wenbopan/Faro-Yi-34B-DPO
name: Open Portuguese LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: HateBR Binary
type: ruanchaves/hatebr
split: test
args:
num_few_shot: 25
metrics:
- type: f1_macro
value: 85.12
name: f1-macro
source:
url: >-
https://huggingface.co/spaces/eduagarcia/open_pt_llm_leaderboard?query=wenbopan/Faro-Yi-34B-DPO
name: Open Portuguese LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: PT Hate Speech Binary
type: hate_speech_portuguese
split: test
args:
num_few_shot: 25
metrics:
- type: f1_macro
value: 68.88
name: f1-macro
source:
url: >-
https://huggingface.co/spaces/eduagarcia/open_pt_llm_leaderboard?query=wenbopan/Faro-Yi-34B-DPO
name: Open Portuguese LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: tweetSentBR
type: eduagarcia/tweetsentbr_fewshot
split: test
args:
num_few_shot: 25
metrics:
- type: f1_macro
value: 72.24
name: f1-macro
source:
url: >-
https://huggingface.co/spaces/eduagarcia/open_pt_llm_leaderboard?query=wenbopan/Faro-Yi-34B-DPO
name: Open Portuguese LLM Leaderboard
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. ...
Open Portuguese LLM Leaderboard Evaluation Results
Detailed results can be found here and on the 🚀 Open Portuguese LLM Leaderboard
| Metric | Value |
|---|---|
| Average | 73.82 |
| ENEM Challenge (No Images) | 73.55 |
| BLUEX (No Images) | 65.79 |
| OAB Exams | 55.85 |
| Assin2 RTE | 92.20 |
| Assin2 STS | 79.78 |
| FaQuAD NLI | 71 |
| HateBR Binary | 85.12 |
| PT Hate Speech Binary | 68.88 |
| tweetSentBR | 72.24 |