Qwen/Qwen2-7B-Instruct
Full fine-tune vs Qwen/Qwen2-7B. 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
44 derivatives checked
- MaziyarPanahi/Qwen2-7B-Instruct-GGUF: Same weights, converted
- Qwen/Qwen2-7B-Instruct-GGUF: Same weights, converted
- legraphista/Qwen2-7B-Instruct-IMat-GGUF: Same weights, converted
- mradermacher/Qwen2-7B-Instruct-i1-GGUF: Quantization
- bartowski/Qwen2-7B-Instruct-GGUF: Same weights, converted
- Qwen/Qwen2-7B-Instruct-AWQ: Quantization
- mradermacher/Qwen2-7B-Instruct-GGUF: Same weights, converted
- Tifa-RP/Tifa-7B-Qwen2-v0.1-GGUF: LoRA-style
- featherless-ai-quants/Qwen-Qwen2-7B-Instruct-GGUF: Quantization
- bartowski/Qwen2-7B-Instruct-deccp-GGUF: Quantization
- QuantFactory/Qwen2-7B-BD-RP-GGUF: Full fine-tune
- qwp4w3hyb/Qwen2-7B-Instruct-iMat-GGUF: Same weights, converted
- AITeamVN/Vi-Qwen2-3B-RAG: Different architecture
- nvidia/VideoITG-8B: Full fine-tune (fewer layers)
- second-state/Qwen2-7B-Instruct-GGUF: Same weights, converted
- QuantFactory/Qwen2-7B-Instruct-GGUF: Quantization
- AITeamVN/Vi-Qwen2-1.5B-RAG: Different architecture
- Qwen/Qwen2-7B-Instruct-GPTQ-Int4: Quantization
- RESMPDEV/Qwen2-Wukong-5B-Instruct-Sliced: Partial fine-tune
- gaianet/Qwen2-7B-Instruct-GGUF: Same weights, converted
- neulab/Pangea-7B-hf: Full fine-tune (fewer layers)
- nisten/qwenv2-7b-inst-imatrix-gguf: Same weights, converted
- neulab/Pangea-7B: Full fine-tune (fewer layers)
- CHE-72-ZLab/Alibaba-Qwen2-7B-Instruct-GGUF: Quantization
- NikolayKozloff/Qwen2-7B-Instruct-Q4_K_M-GGUF: Quantization
- GPT4All-Community/Qwen2-7B-Instruct-GGUF: Quantization
- Qwen/Qwen2-7B-Instruct-GPTQ-Int8: Quantization
- DevQuasar/Qwen2-7B-Instruct-GGUF: Quantization
- yale-nlp/MDCure-Qwen2-7B-Instruct: LoRA-style
- liodon-ai/Qwen2-7B-Instruct-imatrix-GGUF: Quantization
17 other repos hold byte-for-byte identical weight files.