Qwen/Qwen2.5-7B-Instruct
Full fine-tune vs Qwen/Qwen2.5-7B. Fine-tune: 100% of the weights changed, across 28 of 28 layers.
Tensors changed: Embeddings 1/1, Norms 55/57, MLP 84/84, Attention 196/196, lm_head 1/1
566 derivatives checked
- Qwen/Qwen2.5-7B-Instruct-AWQ: Quantization
- Qwen/Qwen2.5-7B-Instruct-GPTQ-Int4: Quantization
- bartowski/Qwen2.5-7B-Instruct-GGUF: Same weights, converted
- unsloth/Qwen2.5-7B-Instruct: Same weights, repacked
- Qwen/Qwen2.5-7B-Instruct-GGUF: Same weights, converted
- MaziyarPanahi/Qwen2.5-7B-Instruct-GGUF: Same weights, converted
- unsloth/Qwen2.5-7B-Instruct-bnb-4bit: Quantization
- unsloth/Qwen2.5-7B-Instruct-unsloth-bnb-4bit: Quantization
- RedHatAI/Qwen2.5-7B-Instruct-FP8-dynamic: Quantization
- lmstudio-community/Qwen2.5-7B-Instruct-GGUF: Quantization
- langfeng01/GiGPO-Qwen2.5-7B-Instruct-WebShop: Full fine-tune
- saidutta69/Qwen2.5-7B-Instruct-heretic: Abliterated
- QuantFactory/Qwen2.5-7B-Instruct-abliterated-v2-GGUF: Quantization
- QuantFactory/Qwen2.5-7B-Instruct-Uncensored-GGUF: Quantization
- paultimothymooney/Qwen2.5-7B-Instruct-Q4_K_M-GGUF: Quantization
- huihui-ai/Qwen2.5-7B-Instruct-abliterated-v2: Abliterated
- richardyoung/Qwen2.5-7B-Instruct-abliterated-GGUF: Quantization
- PKU-ML/G1-7B: Full fine-tune
- liuwenhan/reasonrank-7B: Full fine-tune
- Orion-zhen/Qwen2.5-7B-Instruct-Uncensored: LoRA-style
- Efficient-Large-Model/Fast_dLLM_v2_7B: Full fine-tune
- mradermacher/Qwen2.5-7B-Instruct-i1-GGUF: Quantization
- Qwen/Qwen2.5-7B-Instruct-GPTQ-Int8: Quantization
- lmstudio-community/Qwen2.5-7B-Instruct-MLX-4bit: Quantization
- Mungert/pokee_research_7b-GGUF: Partial fine-tune
- kepom/Qwen2.5-7B-Instruct-abliterated-v2-GGUF: Quantization
- Szymon-Wachelko/GPT-neo-Custom-7B-GGUF: Same weights, converted
- thoughtworks/Qwen2.5-7B-Instruct-Eagle3: Different architecture
- mradermacher/Qwen2.5-7B-Instruct-GGUF: Same weights, converted
- elyza/ELYZA-Shortcut-1.0-Qwen-7B: Full fine-tune
82 other repos hold byte-for-byte identical weight files.