Qwen/Qwen2.5-Math-7B-Instruct
Full fine-tune vs Qwen/Qwen2.5-Math-7B. 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, lm_head 1/1
41 derivatives checked
- Qwen/Qwen2.5-Math-PRM-7B: Partial fine-tune
- bartowski/Qwen2.5-Math-7B-Instruct-GGUF: Same weights, converted
- lmstudio-community/Qwen2.5-Math-7B-Instruct-GGUF: Quantization
- mradermacher/Qwen2.5-Math-7B-Instruct-GGUF: Same weights, converted
- mradermacher/Qwen2.5-Math-7B-Instruct-i1-GGUF: Quantization
- legraphista/Qwen2.5-Math-7B-Instruct-IMat-GGUF: Same weights, converted
- QuantFactory/SuperCorrect-7B-GGUF: Full fine-tune
- second-state/Qwen2.5-Math-7B-Instruct-GGUF: Same weights, converted
- RichardErkhov/BitStarWalkin_-_SuperCorrect-7B-gguf: Full fine-tune
- featherless-ai-quants/Qwen-Qwen2.5-Math-7B-Instruct-GGUF: Quantization
- gaianet/Qwen2.5-Math-7B-Instruct-GGUF: Same weights, converted
- drawais/Qwen2.5-Math-7B-Instruct-AWQ-INT4: Quantization
- DevQuasar-8/Qwen.Qwen2.5-Math-7B-Instruct-GGUF: Same weights, converted
- unsloth/Qwen2.5-Math-7B-Instruct-bnb-4bit: Quantization
- chiakelvin/Qwen2.5-Math-7B-Instruct-Q4_K_L.gguf: Same weights, converted
- Qwen/Qwen2.5-Math-7B-PRM800K: Partial fine-tune
- chiakelvin/Qwen2.5-Math-7B-Instruct-Q5_K_M.gguf: Same weights, converted
- AI-MO/Kimina-Prover-Preview-Distill-7B: Full fine-tune
- mlx-community/Qwen2.5-Math-7B-Instruct-4bit: Quantization
- DuoNeural/Qwen2.5-Math-7B-Instruct-GTAP-v3-IQ3_XXS-GGUF: Quantization
- DuoNeural/Qwen2.5-Math-7B-Instruct-CodeInfused-IQ3_XXS-GGUF: Quantization
- MaziyarPanahi/Qwen2.5-Math-7B-Instruct-GGUF: Same weights, converted
- ZaandaTeika/Qwen2.5-Math-7B-SHARP-Span: Full fine-tune
- tsrikant/vce-specialist-math-7b-gguf: Quantization
- DuoNeural/Qwen2.5-Math-NeuralMath-7B: LoRA-style
- tensorblock/Qwen2.5-Math-7B-Instruct-GGUF: Quantization
- declare-lab/PathFinder-PRM-7B: Full fine-tune
- unsloth/Qwen2.5-Math-7B-Instruct: Same weights, repacked
- BitStarWalkin/SuperCorrect-7B: Full fine-tune
- cehao/Select2Reason-Qwen-7B: Full fine-tune
7 other repos hold byte-for-byte identical weight files.