Qwen/Qwen2-72B-Instruct
Full fine-tune vs Qwen/Qwen2-72B. Fine-tune: 100% of the weights changed, across 80 of 80 layers.
Tensors changed: Embeddings 1/1, MLP 240/240, Attention 560/560, Norms 161/161, lm_head 1/1
19 derivatives checked
- mradermacher/Qwen2-72B-Instruct-i1-GGUF: Quantization
- legraphista/Qwen2-72B-Instruct-IMat-GGUF: Same weights, converted
- qwp4w3hyb/Qwen2-72B-Instruct-iMat-GGUF: Same weights, converted
- featherless-ai-quants/Qwen-Qwen2-72B-Instruct-GGUF: Quantization
- mradermacher/Qwen2-72B-Instruct-GGUF: Quantization
- Qwen/Qwen2-72B-Instruct-AWQ: Quantization
- second-state/Qwen2-72B-Instruct-GGUF: Same weights, converted
- gaianet/Qwen2-72B-Instruct-GGUF: Same weights, converted
- MarsupialAI/MG-FinalMix-72B_iMatrix_GGUF: LoRA-style
- kowlick/Qwen2-72B-Instruct-Q4_K_M-GGUF: Quantization
- Qwen/Qwen2-72B-Instruct-GPTQ-Int8: Quantization
- Qwen/Qwen2-72B-Instruct-GPTQ-Int4: Quantization
- DevQuasar-4/Qwen2-72B-Instruct-GGUF: Quantization
- anthracite-org/magnum-v1-72b: Full fine-tune
- anthracite-org/magnum-v2-72b: Full fine-tune
- dfurman/Qwen2-72B-Orpo-v0.1: LoRA-style
- MaziyarPanahi/calme-2.1-qwen2-72b: LoRA-style
- Undi95/MG-FinalMix-72B: LoRA-style
- Delta-Vector/Mag-Picaro-72B: Full fine-tune
2 other repos hold byte-for-byte identical weight files.