Sao10K/L3.3-70B-Euryale-v2.3
LoRA-style vs meta-llama/Llama-3.3-70B-Instruct. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of down_proj, gate_proj, k_proj, o_proj, q_proj, up_proj, v_proj changed, in 80 of 80 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/1, Norms 0/161, MLP 240/240, Attention 320/320, lm_head 0/1
8 derivatives checked
- bartowski/L3.3-70B-Euryale-v2.3-GGUF: Quantization
- backyardai/L3.3-70B-Euryale-v2.3-GGUF: Same weights, converted
- mradermacher/L3.3-70B-Euryale-v2.3-i1-GGUF: Quantization
- mradermacher/L3.3-70B-Euryale-v2.3-GGUF: Quantization
- meganovaai/L3.3-70B-Euryale-v2.3-AWQ: Quantization
- serpro/L3.3-70B-Euryale-v2.3-mlx-8Bit: Quantization
- Wwayu/L3.3-70B-Euryale-v2.3-mlx-4Bit: Quantization
- tssst/Geranium-105B-v1.0.0: Full fine-tune (fewer layers)