Documented Gensyn base
Hub tags and the model card both identify Gensyn/Qwen2.5-0.5B-Instruct as the base model and fine-tune source.
Open Source Model Profile · 0xibe
This 0xibe model is a 494M-parameter Qwen2 text-generation fine-tune. Its model card identifies Gensyn/Qwen2.5-0.5B-Instruct as the base with TRL and GRPO training.
The model is published by 0xibe as a text-generation fine-tune of Gensyn/Qwen2.5-0.5B-Instruct. The captured configuration identifies Qwen2ForCausalLM with a qwen2 model type, and Safetensors metadata reports 494,032,768 parameters. According to the model card, it was trained using TRL, with GRPO named as the training procedure.
Hub tags and the model card both identify Gensyn/Qwen2.5-0.5B-Instruct as the base model and fine-tune source.
According to the model card, training used TRL with GRPO, the method introduced in DeepSeekMath work.
Captured config identifies Qwen2ForCausalLM and qwen2, with Safetensors metadata reporting 494,032,768 parameters.
The model card documents a transformers text-generation pipeline snippet for running the model.
Source: 0xibe/Qwen2.5-0.5B-Instruct-Gensyn-Swarm-pesty_barky_ibis
Captured: Unknown. Processed: 2026-09-07T19:35:01.776296+00:00.
Model Card for Qwen2.5-0.5B-Instruct-Gensyn-Swarm-pesty_barky_ibis This model is a fine-tuned version of Gensyn/Qwen2.5-0.5B-Instruct . It has been trained using TRL . Quick start from transformers import pipeline question = "If you had a time machine, but could only go to the past or the future once and never return, which would you choose and why?" generator = pipeline( "text-generation" , model= "0xibe/Qwen2.5-0.5B-Instruct-Gensyn-Swarm-pesty_barky_ibis" , device= "cuda" ) output = generator([{ "role" : "user" , "content" : question}], max_new_tokens= 128 , return_full_text= False )[ 0 ] print (output[ "generated_text" ]) Trainin…
F001F002F003F004F005F006F007F008F009F010F011F012F013