ALMA-7B-Ja huggingface.co api & webbigdata ALMA-7B-Ja github AI Model

Introduction of ALMA-7B-Ja

Model Details of ALMA-7B-Ja

New Version has been released.

2024/03/04
webbigdata/C3TR-Adapter
Memory GPU requirement has increased to 8.1 GB. However, it is possible to run it with the free version of Colab and the performance is much improved!

2023/10/21
ALMA-7B-Ja-V2
Overall performance has been raised.

Below is a description of the old version. We urge you to try the newer version above.

webbigdata/ALMA-7B-Ja

ALMA-7B-Ja(13.3GB) is a machine translation model that uses ALMA's learning method to translate Japanese to English.
The original ALMA-7B (26.95GB) supports English and Russian(ru) translation. This model supports Japanese(ja) and English translations instead of Russian.

Like the original model, This model has been verified that it also has a translation ability between the following languages, but if you want the translation function for these languages, it is better to use the original ALMA-13B model .

German(de) and English(en)
Chinese(zh) and English(en)
Icelandic(is) and English(en)
Czech(cs) and English(en)

Translating from English (en→xx) BLEU/COMET

Models	de	cs	is	zh	ru/jp	Avg.
NLLB-54B	34.50/86.45	37.60/90.15	24.15/81.76	27.38/78.91	30.96/87.92	30.92/85.04
GPT-3.5-D	31.80/85.61	31.30/88.57	15.90/76.28	38.30/85.76	27.50/86.74	28.96/84.59
ALMA-7B(Original)	30.31/85.59	29.88/89.10	25.71/85.52	36.87/85.11	27.13/86.98	29.89/86.49
ALMA-7B-Ja(Ours)	23.70/82.04	18.58/81.36	12.20/71.59	29.06/82.45	14.82/85.40	19.67/80.57

Translating to English (xx→en) BLEU/COMET

Models	de	cs	is	zh	ru/jp	Avg.
NLLB-54B	26.89/78.94	39.11/80.13	23.09/71.66	16.56/70.70	39.11/81.88	28.95/76.66
GPT-3.5-D	30.90/84.79	44.50/86.16	31.90/82.13	25.00/81.62	38.50/84.80	34.16/83.90
ALMA-7B(Original)	30.26/84.00	43.91/85.86	35.97/86.03	23.75/79.85	39.37/84.58	34.55/84.02
ALMA-7B-Ja(Ours)	26.41/83.13	34.39/83.50	24.77/81.12	20.60/78.54	15.57/78.61	24.35/81.76

Sample Code For Free Colab

Other Version

webbigdata-ALMA-7B-Ja-gguf

mmnga made llama.cpp(gguf) version webbigdata-ALMA-7B-Ja-gguf . Thank you!
llama.cpp is a tool used primarily on Macs, and gguf is its latest version format. It can be used without gpu.

ALMA-7B-Ja-gguf Free Colab sample

ALMA-7B-Ja-GPTQ-Ja-En

GPTQ is quantized(reduce the size of the model) method and ALMA-7B-Ja-GPTQ has GPTQ quantized version that reduces model size(3.9GB) and memory usage.
But the performance is probably lower. And translation ability for languages other than Japanese and English has deteriorated significantly.

Sample Code For Free Colab webbigdata/ALMA-7B-Ja-GPTQ-Ja-En

If you want to translate the entire file at once, try Colab below.
ALMA_7B_Ja_GPTQ_Ja_En_batch_translation_sample

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 their paper .

@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}
}

about this work

This work was done by : webbigdata .

Runs of webbigdata ALMA-7B-Ja on huggingface.co

Total runs

24-hour runs

-1

3-day runs

-2

7-day runs

30-day runs

More Information About ALMA-7B-Ja huggingface.co Model

ALMA-7B-Ja huggingface.co

ALMA-7B-Ja huggingface.co is an AI model on huggingface.co that provides ALMA-7B-Ja's model effect (), which can be used instantly with this webbigdata ALMA-7B-Ja model. huggingface.co supports a free trial of the ALMA-7B-Ja model, and also provides paid use of the ALMA-7B-Ja. Support call ALMA-7B-Ja model through api, including Node.js, Python, http.

ALMA-7B-Ja huggingface.co Url

https://huggingface.co/webbigdata/ALMA-7B-Ja

webbigdata ALMA-7B-Ja online free

ALMA-7B-Ja huggingface.co is an online trial and call api platform, which integrates ALMA-7B-Ja's modeling effects, including api services, and provides a free online trial of ALMA-7B-Ja, you can try ALMA-7B-Ja online for free by clicking the link below.

webbigdata ALMA-7B-Ja online free url in huggingface.co:

https://huggingface.co/webbigdata/ALMA-7B-Ja

ALMA-7B-Ja install

ALMA-7B-Ja is an open source model from GitHub that offers a free installation service, and any user can find ALMA-7B-Ja on GitHub to install. At the same time, huggingface.co provides the effect of ALMA-7B-Ja install, users can directly use ALMA-7B-Ja installed effect in huggingface.co for debugging and trial. It also supports api for free installation.