manindra18/Qwen3-8B is an 8 billion parameter causal language model fine-tuned from Qwen/Qwen3-8B. This model was fine-tuned using Tinker and tinker-cookbook, resulting in a merged model format. It is suitable for general text generation tasks, leveraging the base Qwen3-8B architecture.
Loading preview...
Overview
manindra18/Qwen3-8B is an 8 billion parameter language model, fine-tuned from the robust Qwen/Qwen3-8B base model. The fine-tuning process utilized the Tinker framework and the associated tinker-cookbook, resulting in a merged model format.
Key Characteristics
- Base Model: Qwen/Qwen3-8B, a well-regarded open-source large language model.
- Parameter Count: 8 billion parameters, offering a balance between performance and computational efficiency.
- Context Length: Supports a context window of 32768 tokens, enabling processing of longer inputs and generating more coherent, extended outputs.
- Fine-tuning Method: Fine-tuned using Tinker, a specialized framework for model adaptation.
Usage
This model can be easily loaded and used with the Hugging Face transformers library for various natural language processing tasks, particularly those involving text generation and understanding. Its merged format ensures straightforward deployment.
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("manindra18/Qwen3-8B")Framework Versions
The fine-tuning and development environment included:
tinker-cookbook: 0.1.0transformers: 5.1.0torch: 2.10.0