# Low-resource training

ContentV is an efficient video generation model framework that achieves high-quality video generation with limited computing resources through a minimalist architecture, multi-stage training strategy, and cost-effective reinforcement learning framework.

Fine-tuned version of Orpheus 3B TTS model for Bengali, trained with 955 audio samples, suitable for experimental Bengali speech synthesis

A Thai female voice text-to-speech (TTS) model based on the VITS architecture, supporting high-quality Thai speech synthesis.

AuroraCap is a multimodal large language model for image and video captioning, focusing on efficient and detailed video caption generation.

An upgraded version of the SmolLM-135M micro language model with 0.18B parameters, suitable for training scenarios, featuring excellent inference speed and coherence performance

BitNet b1.58 3B is a 1-bit quantized large language model trained on 100 billion tokens from the RedPajama dataset, significantly reducing computational resource requirements while maintaining performance.

This is a small GPT-2 model pre-trained on the TinyStories V2 dataset, featuring 3M trainable parameters and demonstrating good text generation coherence.

A 33M-parameter small language model trained on the TinyStories dataset, specifically designed for generating children's stories

TinyStories-1M is a small language model trained on the TinyStories dataset, specifically designed to generate simple stories suitable for children's reading.

Tiny ImageNet is a small-scale image classification dataset designed for benchmarking and model training in computer vision tasks.

Open-source Chinese conversational large language model optimized with instruction fine-tuning, specializing in Chinese cultural tasks, with 1.4B/2.6B parameters

A speech recognition model fine-tuned on the Japanese CommonVoice dataset based on OpenAI's whisper-large-v2 model, trained for 5000 steps with a word error rate of 0.7449

A text summarization generation model fine-tuned based on google/mt5-small, supporting multilingual text summarization tasks.

This is a sequence-to-sequence model based on the T5-small architecture, specifically designed for translating Hindi-English mixed text (hi_en) into standard English (en).

This model is a speech recognition model fine-tuned on the LibriSpeech dataset based on facebook/wav2vec2-base, suitable for English speech-to-text tasks.

This model is a fine-tuned speech processing model based on facebook/wav2vec2-base, with no specific use case explicitly stated

A speech recognition model fine-tuned based on facebook/wav2vec2-base, trained on 10ms masked audio tasks

A Portuguese language model fine-tuned based on GPT-2 small model, trained on Portuguese Wikipedia, supporting NLP tasks like text generation

BabyBERTa is a lightweight version of RoBERTa, trained on child-directed input and specifically designed for language acquisition research.

This model is trained using the WECHSEL method, achieving cross-lingual transfer of monolingual language models through efficient initialization of subword embeddings, specifically optimized for German.

GPT2 model pre-trained for Italian language, 117M parameters, trained on Italian Wikipedia and ItWac corpus