Aletheia-Bench/BatchOnline-GRPO-1.5B
Partial fine-tune vs deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B. Partial change: 87% of the weights changed (attention, embeddings, mlp, norms), in 28 of 28 layers.
Tensors changed: Embeddings 1/1, Attention 196/196, MLP 84/84, Norms 32/57, lm_head 0/1
2 derivatives checked
- mradermacher/BatchOnline-GRPO-1.5B-i1-GGUF: Quantization
- mradermacher/BatchOnline-GRPO-1.5B-GGUF: Same weights, converted
1 other repos hold byte-for-byte identical weight files.