Qwen/Qwen2.5-Math-1.5B
Full fine-tune vs Qwen/Qwen2.5-1.5B. Fine-tune: 100% of the weights changed, across 28 of 28 layers.
Tensors changed: Embeddings 1/1, Norms 57/57, MLP 84/84, Attention 196/196
43 derivatives checked
- Qwen/Qwen2.5-Math-1.5B-Instruct: Full fine-tune
- WeiboAI/VibeThinker-1.5B: Full fine-tune
- unsloth/Qwen2.5-Math-1.5B: Identical copy
- Mungert/VibeThinker-1.5B-GGUF: Full fine-tune
- Mungert/OpenMath-Nemotron-1.5B-GGUF: Full fine-tune
- mradermacher/Qwen2.5-Math-1.5B-i1-GGUF: Quantization
- QuantFactory/Qwen2.5-Math-1.5B-Instruct-GGUF: Full fine-tune
- mradermacher/Qwen2.5-Math-1.5B-GGUF: Same weights, converted
- featherless-ai-quants/Qwen-Qwen2.5-Math-1.5B-GGUF: Quantization
- nvidia/OpenMath-Nemotron-1.5B: Full fine-tune
- zbeeb/Qwen2.5-Math-1.5B-GRPO-Staleness-2: Full fine-tune
- sail/Qwen2.5-Math-1.5B-Oat-Zero: Full fine-tune
- unsloth/Qwen2.5-Math-1.5B-bnb-4bit: Quantization
- TIGER-Lab/general-verifier: Full fine-tune
- zbeeb/Qwen2.5-Math-1.5B-GRPO-Staleness-6: Full fine-tune
- zbeeb/Qwen2.5-Math-1.5B-Base-GRPO: Full fine-tune
- zbeeb/Qwen2.5-Math-1.5B-GRPO-Staleness-8: Full fine-tune
- JaeWooShin/qwen2.5-math-1.5b-fewgt-u1-n256-inverted-lam1-s0: Full fine-tune
- ypwang61/One-Shot-RLVR-Qwen2.5-Math-1.5B-pi1: Full fine-tune
- menik1126/OVD-Math-1-data-FR-Step300-Historical: Partial fine-tune
- menik1126/OVD-Math-1-data-FR-Step500-Historical: Partial fine-tune
- menik1126/OVD-Math-1-data-FR-Step600-Historical: Partial fine-tune
- menik1126/OVD-Math-1-data-HES-Step500-Historical: Partial fine-tune
- menik1126/OVD-Math-1-data-FR-Step800-Historical: Partial fine-tune
- menik1126/OVD-Math-1-data-RS-Step300-Historical: Partial fine-tune
- menik1126/OVD-Math-1-data-HES-Step300-Historical: Partial fine-tune
- menik1126/OVD-Math-1-data-HES-Step800-Historical: Partial fine-tune
- menik1126/OVD-Math-1-data-RS-Step800-Historical: Partial fine-tune
- menik1126/OVD-Math-1-data-HES-Step600-Historical: Partial fine-tune
- menik1126/OVD-Math-1-data-RS-Step500-Historical: Partial fine-tune
17 other repos hold byte-for-byte identical weight files.