doing something
https://huggingface.co/spaces/Yehor/en-uk-translator Now you can play with the model I will update the checkpoint sometimes #nlp #llm
Updated to v5 (beta)
It was trained on the half of dataset, now translations are better than before
It was trained on the half of dataset, now translations are better than before
doing something
Updated to v4.5 (beta) Mother model also gives some multilingual capabilities #nlp
Also, an example of this behaviour
doing something
https://huggingface.co/spaces/Yehor/en-uk-translator Now you can play with the model I will update the checkpoint sometimes #nlp #llm
This model in quantized variants should be usable on mobile devices
Forwarded from Feed-Master
Amazon
Introducing Amazon S3 Vectors: First cloud storage with native vector support at scale (preview) | Amazon Web Services
Amazon S3 Vectors is a new cloud object store that provides native support for storing and querying vectors at massive scale, offering up to 90% cost reduction compared to conventional approaches while seamlessly integrating with Amazon Bedrock Knowledge…
❤🔥1
doing something
https://huggingface.co/spaces/Yehor/en-uk-translator Now you can play with the model I will update the checkpoint sometimes #nlp #llm
Media is too big
VIEW IN TELEGRAM
A new lightweight model to do machine translation from English to Ukrainian using recently published LFM2 model
Facts:
- Fine-tuned with 40M samples (filtered by quality metric) from ~53.5M for 1.4 epochs
- 354M params
- Requires 1 GB of RAM to run with bf16
- BLEU on FLORES-200: 27.24
- In video I've used RTX 3090 Ti for inference
- Tokens per second: 229.93 (bs=1), 1664.40 (bs=10), 8392.48 (bs=64)
- License: lfm1.0 (free usage for commercial if you have <$10m revenue)
Links:
- Model: https://huggingface.co/Yehor/kulyk-en-uk
- Demo: https://huggingface.co/spaces/Yehor/en-uk-translator
Acknowledgements:
- Serhiy Stetskovych for providing compute to train this model
- lang-uk members for their compilation of different MT datasets
#nlp #llm
Facts:
- Fine-tuned with 40M samples (filtered by quality metric) from ~53.5M for 1.4 epochs
- 354M params
- Requires 1 GB of RAM to run with bf16
- BLEU on FLORES-200: 27.24
- In video I've used RTX 3090 Ti for inference
- Tokens per second: 229.93 (bs=1), 1664.40 (bs=10), 8392.48 (bs=64)
- License: lfm1.0 (free usage for commercial if you have <$10m revenue)
Links:
- Model: https://huggingface.co/Yehor/kulyk-en-uk
- Demo: https://huggingface.co/spaces/Yehor/en-uk-translator
Acknowledgements:
- Serhiy Stetskovych for providing compute to train this model
- lang-uk members for their compilation of different MT datasets
#nlp #llm
🔥4
doing something
A new lightweight model to do machine translation from English to Ukrainian using recently published LFM2 model Facts: - Fine-tuned with 40M samples (filtered by quality metric) from ~53.5M for 1.4 epochs - 354M params - Requires 1 GB of RAM to run with…
It learned russian as well
Idk why and how
Idk why and how
😢7
doing something
A new lightweight model to do machine translation from English to Ukrainian using recently published LFM2 model Facts: - Fine-tuned with 40M samples (filtered by quality metric) from ~53.5M for 1.4 epochs - 354M params - Requires 1 GB of RAM to run with…
LiquidAI has a space where you can load entire model into your browser and perform generation
It's based on GGUF version of the model
I plan to make this type of Translator as well
It's based on GGUF version of the model
I plan to make this type of Translator as well
🔥1
doing something
A new lightweight model to do machine translation from English to Ukrainian using recently published LFM2 model Facts: - Fine-tuned with 40M samples (filtered by quality metric) from ~53.5M for 1.4 epochs - 354M params - Requires 1 GB of RAM to run with…
huggingface.co
🇺🇦 UK-EN Translator - a Hugging Face Space by Yehor
This application translates text, audio, and images from Ukrainian to English. Users can input text, upload audio files, or upload images, and the app will provide the translated content in English.