model-index, library_name, license, license_name, license_link, language, pipeline_tag, tags, datasets, base_model

model-index	library_name	license	license_name	license_link	language	pipeline_tag	tags	datasets	base_model
name results Kulyk-UK-EN task dataset metrics type text-generation type name split facebook/flores FLORES devtest type value name bleu 36.27 BLEU	transformers	other	lfm1.0	LICENSE	en uk	text-generation	liquid lfm2 edge	lang-uk/FiftyFiveShades	LiquidAI/LFM2-350M

A lightweight model to do machine translation from Ukrainian to English based on recently published LFM2 model. Use demo to test it.

Also, there's another model: kulyk-en-uk

Run with Docker (CPU):

docker run -p 3000:3000 --rm ghcr.io/egorsmkv/kulyk-rust:latest

Run using Apptainer (CUDA):

1. Run it using shell:

apptainer shell --nv ./kulyk.sif

Apptainer> /opt/entrypoints/kulyk --verbose --n-len 1024 --model-path-ue /project/models/kulyk-uk-en.gguf --model-path-eu /project/models/kulyk-en-uk.gguf

2. Run it as a webservice:

apptainer instance start --nv ./kulyk.sif kulyk-ws

# go to http://localhost:3000

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: 36.27
• Tokens per second: 229.93 (bs=1), 1664.40 (bs=10), 8392.48 (bs=64)
• License: lfm1.0

Info:

• Model name is inherited from name of Sergiy Kulyk who was chargé d'affaires of Ukraine in the United States

Training Info:

• Learning Rate: 3e-5
• Learning Rate scheduler type: cosine
• Warmup Ratio: 0.05
• Max length: 2048
• Batch Size: 10
• packed=True
• Sentences <= 1000 chars
• Gradient accumulation steps: 4
• Used Flash Attention 2
• Time for epoch: 32 hours
• 2 cards of NVIDIA RTX 3090 Ti (24G)
• accelerate with DeepSpeed
• Memory usage: 22.212GB-22.458GB
• torch 2.7.1

Acknowledgements:

• Dmytro Chaplynskyi for providing compute to train this model
• lang-uk members for their compilation of different MT datasets