|
--- |
|
license: mit |
|
model-index: |
|
- name: ALMA-7B |
|
results: |
|
- task: |
|
type: text-generation |
|
name: Text Generation |
|
dataset: |
|
name: AI2 Reasoning Challenge (25-Shot) |
|
type: ai2_arc |
|
config: ARC-Challenge |
|
split: test |
|
args: |
|
num_few_shot: 25 |
|
metrics: |
|
- type: acc_norm |
|
value: 50.34 |
|
name: normalized accuracy |
|
source: |
|
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=haoranxu/ALMA-7B |
|
name: Open LLM Leaderboard |
|
- task: |
|
type: text-generation |
|
name: Text Generation |
|
dataset: |
|
name: HellaSwag (10-Shot) |
|
type: hellaswag |
|
split: validation |
|
args: |
|
num_few_shot: 10 |
|
metrics: |
|
- type: acc_norm |
|
value: 75.5 |
|
name: normalized accuracy |
|
source: |
|
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=haoranxu/ALMA-7B |
|
name: Open LLM Leaderboard |
|
- task: |
|
type: text-generation |
|
name: Text Generation |
|
dataset: |
|
name: MMLU (5-Shot) |
|
type: cais/mmlu |
|
config: all |
|
split: test |
|
args: |
|
num_few_shot: 5 |
|
metrics: |
|
- type: acc |
|
value: 38.04 |
|
name: accuracy |
|
source: |
|
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=haoranxu/ALMA-7B |
|
name: Open LLM Leaderboard |
|
- task: |
|
type: text-generation |
|
name: Text Generation |
|
dataset: |
|
name: TruthfulQA (0-shot) |
|
type: truthful_qa |
|
config: multiple_choice |
|
split: validation |
|
args: |
|
num_few_shot: 0 |
|
metrics: |
|
- type: mc2 |
|
value: 35.64 |
|
source: |
|
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=haoranxu/ALMA-7B |
|
name: Open LLM Leaderboard |
|
- task: |
|
type: text-generation |
|
name: Text Generation |
|
dataset: |
|
name: Winogrande (5-shot) |
|
type: winogrande |
|
config: winogrande_xl |
|
split: validation |
|
args: |
|
num_few_shot: 5 |
|
metrics: |
|
- type: acc |
|
value: 72.38 |
|
name: accuracy |
|
source: |
|
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=haoranxu/ALMA-7B |
|
name: Open LLM Leaderboard |
|
- task: |
|
type: text-generation |
|
name: Text Generation |
|
dataset: |
|
name: GSM8k (5-shot) |
|
type: gsm8k |
|
config: main |
|
split: test |
|
args: |
|
num_few_shot: 5 |
|
metrics: |
|
- type: acc |
|
value: 0.0 |
|
name: accuracy |
|
source: |
|
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=haoranxu/ALMA-7B |
|
name: Open LLM Leaderboard |
|
--- |
|
**ALMA** (**A**dvanced **L**anguage **M**odel-based tr**A**nslator) is an LLM-based translation model, which adopts a new translation model paradigm: it begins with fine-tuning on monolingual data and is further optimized using high-quality parallel data. This two-step fine-tuning process ensures strong translation performance. |
|
Please find more details in our [paper](https://arxiv.org/abs/2309.11674). |
|
``` |
|
@misc{xu2023paradigm, |
|
title={A Paradigm Shift in Machine Translation: Boosting Translation Performance of Large Language Models}, |
|
author={Haoran Xu and Young Jin Kim and Amr Sharaf and Hany Hassan Awadalla}, |
|
year={2023}, |
|
eprint={2309.11674}, |
|
archivePrefix={arXiv}, |
|
primaryClass={cs.CL} |
|
} |
|
``` |
|
**[ALMA-R](https://arxiv.org/abs/2401.08417) (NEW!) is released now!** ALMA-R builds upon ALMA models, with further LoRA fine-tuning with our proposed **Contrastive Preference Optimization (CPO)** as opposed to the Supervised Fine-tuning used in ALMA. CPO fine-tuning requires our [triplet preference data](https://huggingface.co/datasets/haoranxu/ALMA-R-Preference) for preference learning. ALMA-R now can matches or even exceeds GPT-4 or WMT winners! |
|
|
|
|
|
``` |
|
@misc{xu2024contrastive, |
|
title={Contrastive Preference Optimization: Pushing the Boundaries of LLM Performance in Machine Translation}, |
|
author={Haoran Xu and Amr Sharaf and Yunmo Chen and Weiting Tan and Lingfeng Shen and Benjamin Van Durme and Kenton Murray and Young Jin Kim}, |
|
year={2024}, |
|
eprint={2401.08417}, |
|
archivePrefix={arXiv}, |
|
primaryClass={cs.CL} |
|
} |
|
``` |
|
We release six translation models presented in the paper: |
|
- **ALMA-7B**: Full-weight Fine-tune LLaMA-2-7B on 20B monolingual tokens and then **Full-weight** fine-tune on human-written parallel data |
|
- **ALMA-7B-LoRA**: Full-weight Fine-tune LLaMA-2-7B on 20B monolingual tokens and then **LoRA** fine-tune on human-written parallel data |
|
- **ALMA-7B-R (NEW!)**: Further LoRA fine-tuning upon ALMA-7B-LoRA with contrastive preference optimization. |
|
- **ALMA-13B**: Full-weight Fine-tune LLaMA-2-7B on 12B monolingual tokens and then **Full-weight** fine-tune on human-written parallel data |
|
- **ALMA-13B-LoRA** (Our best system): Full-weight Fine-tune LLaMA-2-7B on 12B monolingual tokens and then **LoRA** fine-tune on human-written parallel data |
|
- **ALMA-13B-R (NEW!)**: Further LoRA fine-tuning upon ALMA-13B-LoRA with contrastive preference optimization. |
|
|
|
Model checkpoints are released at huggingface: |
|
| Models | Base Model Link | LoRA Link | |
|
|:-------------:|:---------------:|:---------:| |
|
| ALMA-7B | [haoranxu/ALMA-7B](https://huggingface.co/haoranxu/ALMA-7B) | - | |
|
| ALMA-7B-LoRA | [haoranxu/ALMA-7B-Pretrain](https://huggingface.co/haoranxu/ALMA-7B-Pretrain) | [haoranxu/ALMA-7B-Pretrain-LoRA](https://huggingface.co/haoranxu/ALMA-7B-Pretrain-LoRA) | |
|
| **ALMA-7B-R (NEW!)** | [haoranxu/ALMA-7B-R (LoRA merged)](https://huggingface.co/haoranxu/ALMA-7B-R) | - | |
|
| ALMA-13B | [haoranxu/ALMA-13B](https://huggingface.co/haoranxu/ALMA-13B) | - | |
|
| ALMA-13B-LoRA | [haoranxu/ALMA-13B-Pretrain](https://huggingface.co/haoranxu/ALMA-13B-Pretrain) | [haoranxu/ALMA-13B-Pretrain-LoRA](https://huggingface.co/haoranxu/ALMA-13B-Pretrain-LoRA) | |
|
| **ALMA-13B-R (NEW!)** | [haoranxu/ALMA-13B-R (LoRA merged)](https://huggingface.co/haoranxu/ALMA-13B-R) | - | |
|
|
|
**Note that `ALMA-7B-Pretrain` and `ALMA-13B-Pretrain` are NOT translation models. They only experience stage 1 monolingual fine-tuning (20B tokens for the 7B model and 12B tokens for the 13B model), and should be utilized in conjunction with their LoRA models.** |
|
|
|
Datasets used by ALMA and ALMA-R are also released at huggingface now (NEW!) |
|
| Datasets | Train / Validation| Test | |
|
|:-------------:|:---------------:|:---------:| |
|
| Human-Written Parallel Data (ALMA) | [train and validation](https://huggingface.co/datasets/haoranxu/ALMA-Human-Parallel) | [WMT'22](https://huggingface.co/datasets/haoranxu/WMT22-Test) | |
|
| Triplet Preference Data | [train](https://huggingface.co/datasets/haoranxu/ALMA-R-Preference) | [WMT'22](https://huggingface.co/datasets/haoranxu/WMT22-Test) and [WMT'23](https://huggingface.co/datasets/haoranxu/WMT23-Test) | |
|
|
|
A quick start to use system ALMA-13B-LoRA for translation. An example of translating "我爱机器翻译。" into English: |
|
``` |
|
import torch |
|
from peft import PeftModel |
|
from transformers import AutoModelForCausalLM |
|
from transformers import LlamaTokenizer |
|
|
|
# Load base model and LoRA weights |
|
model = AutoModelForCausalLM.from_pretrained("haoranxu/ALMA-13B-Pretrain", torch_dtype=torch.float16, device_map="auto") |
|
model = PeftModel.from_pretrained(model, "haoranxu/ALMA-13B-Pretrain-LoRA") |
|
tokenizer = LlamaTokenizer.from_pretrained("haoranxu/ALMA-13B-Pretrain", padding_side='left') |
|
|
|
# Add the source setence into the prompt template |
|
prompt="Translate this from Chinese to English:\nChinese: 我爱机器翻译。\nEnglish:" |
|
input_ids = tokenizer(prompt, return_tensors="pt", padding=True, max_length=40, truncation=True).input_ids.cuda() |
|
|
|
# Translation |
|
with torch.no_grad(): |
|
generated_ids = model.generate(input_ids=input_ids, num_beams=5, max_new_tokens=20, do_sample=True, temperature=0.6, top_p=0.9) |
|
outputs = tokenizer.batch_decode(generated_ids, skip_special_tokens=True) |
|
print(outputs) |
|
``` |
|
|
|
Please find more details in our [GitHub repository](https://github.com/fe1ixxu/ALMA) |
|
# [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard) |
|
Detailed results can be found [here](https://huggingface.co/datasets/open-llm-leaderboard/details_haoranxu__ALMA-7B) |
|
|
|
| Metric |Value| |
|
|---------------------------------|----:| |
|
|Avg. |45.32| |
|
|AI2 Reasoning Challenge (25-Shot)|50.34| |
|
|HellaSwag (10-Shot) |75.50| |
|
|MMLU (5-Shot) |38.04| |
|
|TruthfulQA (0-shot) |35.64| |
|
|Winogrande (5-shot) |72.38| |
|
|GSM8k (5-shot) | 0.00| |
|
|
|
|
