# English Text Generation

UniReason-Qwen3-14B-RL is a quantized model applicable to multiple domains, especially proficient in text generation and mathematical reasoning tasks.

This model is converted from canopylabs/orpheus-3b-0.1-ft into GGUF format, suitable for text generation tasks.

A fine-tuned large language model based on mlabonne/gemma-3-27b-it-abliterated, trained using the Unsloth acceleration framework and Huggingface's TRL library, achieving a 2x efficiency improvement.

Llama model trained with Unsloth and Huggingface TRL library, achieving 2x faster inference speed

Llama model trained with Unsloth and Huggingface TRL library, achieving 2x inference speed improvement

RWKV-7 model using Flash Linear Attention format, trained on the Pile dataset, supporting English text generation tasks.

The RWKV-7 model in flash-linear attention format, supporting English text generation tasks.

A model fine-tuned based on unsloth/Llama-3.2-11B-Vision-Instruct, trained using Unsloth and Huggingface's TRL library, achieving a 2x speedup.

Doge is a sequence transformation model that employs dynamic masked attention mechanisms, capable of state transitions using either multi-layer perceptrons or cross-domain mixture of experts.

An instruction-following model fine-tuned based on unsloth/Qwen2-7B-Instruct, trained using Unsloth and TRL libraries with 2x speed improvement

NanoLM-1B-Instruct-v1.1 is a small instruction-tuned language model with 1 billion parameters, supporting multi-domain English text generation tasks.

A 4-bit quantized version based on Meta-Llama-3.1-8B-Instruct, trained using Unsloth and Huggingface TRL libraries, achieving 2x speed improvement.

An 8.9 billion parameter language model developed in the third iteration using self-play preference optimization, starting from google/gemma-2-9b-it and fine-tuned with the UltraFeedback dataset

Athena-70B-L3 is a large language model with 70B parameters, supporting English text generation tasks and utilizing parameter-efficient fine-tuning techniques.

Gemma is a series of lightweight open large language models launched by Google, built on the same technology used to create Gemini models, suitable for various text generation tasks.

Chewy-Lemon-Cookie-11B is an 11B-parameter large language model based on the Mistral architecture, focusing on text generation and role-playing tasks.

Shot2Story-20K is an image-to-text generation model capable of converting input images into coherent textual descriptions or stories.

MythoMax L2 13b is a large language model based on the Q4_K_M quantized version, suitable for text generation tasks.

The instruction-tuned version of Meta Llama 3 with 8B parameters, optimized for dialogue scenarios, demonstrating excellent helpfulness and safety performance.

A 141-billion parameter Mixture of Experts (MoE) model fine-tuned from Mixtral-8x22B-v0.1, with 35 billion active parameters, primarily designed for English text generation tasks

RecurrentGemma is an open language model family developed by Google based on a novel recurrent architecture, offering both pre-trained and instruction-tuned versions suitable for various text generation tasks.

Gemma is a lightweight open model series launched by Google, built on the same technology as Gemini, suitable for various text generation tasks.

Ministral is a small-scale language model series based on the Mistral architecture, with a parameter size of 3 billion, primarily designed for English text generation tasks.

A large language model based on Google's Gemma 7B, fine-tuned using the Zephyr SFT recipe, primarily for text generation tasks

LiteLlama is a streamlined version of Meta AI's LLaMa 2, featuring only 460 million parameters and trained on 1 trillion tokens as an open-source language model

A language model scaled down from Mistral 7B to 248 million parameters, designed for text generation tasks and suitable for downstream task fine-tuning.

Mythalion 13B is a 13B-parameter large language model developed by PygmalionAI, based on the Llama architecture, specializing in text generation and instruction-following tasks.

Llama 2 is a 7-billion-parameter pre-trained generative text model developed by Meta, part of the open-source large language model series

A 111M parameter model in the Cerebras-GPT series, adopting GPT-3 style architecture, trained on The Pile dataset, achieving compute-optimal performance following Chinchilla scaling laws.

Pythia-1B is a language model specialized for interpretability research developed by EleutherAI, belonging to the 1-billion-parameter version in the Pythia suite, trained on The Pile dataset.

Pythia-12B is the largest model in EleutherAI's scalable language model suite, with 12 billion parameters, specifically designed to advance scientific research on large language models

An English event reasoning model based on the T5 architecture, used to analyze event prerequisites, effects, intentions, and reactions

A text generation model fine-tuned on the samsum dialogue summarization dataset based on Google's flan-t5-base model, excelling in dialogue summarization tasks

Pythia-6.9B is a large-scale language model developed by EleutherAI, part of the Pythia scalable suite, specifically designed to facilitate interpretability research.

Pythia-1B Deduplicated is a language model developed by EleutherAI specifically for interpretability research, trained on the deduplicated Pile dataset using Transformer architecture with 1 billion parameters

Pythia is a series of causal language models developed by EleutherAI, specifically designed for interpretability research. It includes 8 model sizes ranging from 70 million to 12 billion parameters, providing 154 training checkpoints.

Pythia-1.4B is a 1.2 billion parameter causal language model developed by EleutherAI, part of the Pythia scale suite, specifically designed for interpretability research.

OPT is an open-source large language model series launched by Meta AI, with parameter scales ranging from 125 million to 175 billion, aimed at promoting open research in large-scale language models.

OPT is an open-source large language model series launched by Meta AI, benchmarked against the GPT-3 architecture, aimed at promoting reproducibility and societal impact discussions in large model research.

T5 Efficient XL is a deep narrow variant of Google's T5 model, focusing on improving downstream task performance by increasing model depth rather than width.