hjsh/qwen2.5_math_1.5b_grpo_prob_adv_scaled_ratio_w_o_kl_step300

Name: hjsh/qwen2.5_math_1.5b_grpo_prob_adv_scaled_ratio_w_o_kl_step300 API
Brand: Featherless.ai
Price: 10.00 USD
Availability: InStock
Author: hjsh

Hugging Face

TEXT GENERATIONConcurrency Cost:1Model Size:1.5BQuant:BF16Ctx Length:32kPublished:May 16, 2026Architecture:Transformer Warm

Loading preview...

Finetunes

1 models

Full Model Card (README)