Qwen/Qwen2.5-Coder-1.5B-Instruct
Full fine-tune vs Qwen/Qwen2.5-Coder-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
110 derivatives checked
- Qwen/Qwen2.5-Coder-1.5B-Instruct-GGUF: Full fine-tune
- ThorOdinson246/nl2sh-1.5b-Q4_K_M: Full fine-tune
- bartowski/Qwen2.5-Coder-1.5B-Instruct-GGUF: Same weights, converted
- unsloth/Qwen2.5-Coder-1.5B-Instruct-bnb-4bit: Quantization
- unsloth/Qwen2.5-Coder-1.5B-Instruct: Identical copy
- lmstudio-community/Qwen2.5-Coder-1.5B-Instruct-GGUF: Quantization
- lmstudio-community/Qwen2.5-Coder-1.5B-Instruct-MLX-8bit: Quantization
- alphaduriendur/Qwen2.5-Coder-1.5B-Instruct-Q4_K_M-GGUF: Quantization
- unsloth/Qwen2.5-Coder-1.5B-Instruct-GGUF: Same weights, converted
- Qwen/Qwen2.5-Coder-1.5B-Instruct-AWQ: Quantization
- unsloth/Qwen2.5-Coder-1.5B-Instruct-128K-GGUF: Same weights, converted
- ggml-org/Qwen2.5-Coder-1.5B-Instruct-Q8_0-GGUF: Quantization
- bartowski/Qwen2.5.1-Coder-1.5B-Instruct-GGUF: Full fine-tune
- ajayk007/Qwen2.5-Coder-1.5B-Shellsmith: Quantization
- saidutta69/Qwen2.5-Coder-1.5B-Instruct-heretic: Abliterated
- apto-as/Qwen2.5-Coder-1.5B-Instruct-Q5_K_M-GGUF: Quantization
- lmstudio-community/Qwen2.5-Coder-1.5B-Instruct-MLX-4bit: Quantization
- thirdeyeai/Qwen2.5-1.5B-Instruct-uncensored: Full fine-tune
- second-state/Qwen2.5-Coder-1.5B-Instruct-GGUF: Same weights, converted
- legraphista/Qwen2.5-Coder-1.5B-Instruct-IMat-GGUF: Same weights, converted
- mradermacher/Qwen2.5-Coder-1.5B-Instruct-i1-GGUF: Quantization
- sheldonrobinson/Qwen2.5-Coder-1.5B-Instruct-Q4_K_M-GGUF: Quantization
- temper-ai/Temper-1-1.5B: Full fine-tune
- gaianet/Qwen2.5-Coder-1.5B-Instruct-GGUF: Same weights, converted
- mradermacher/Qwen2.5-Coder-1.5B-Instruct-GGUF: Same weights, converted
- Nanthasit/sakthai-coder-1.5b: Full fine-tune
- axieyangb/Vesper-Coder-1.5B: LoRA-style
- koshuro/Qwen2.5-Hermes-Instruct-1.5B: LoRA-style
- BlazingCustoms/pybytecode-v3-1.5b: LoRA-style
- minsore/pepper-1-preview: LoRA-style
24 other repos hold byte-for-byte identical weight files.