EVA-UNIT-01/EVA-LLaMA-3.33-70B-v0.0
Partial fine-tune vs meta-llama/Llama-3.3-70B-Instruct. Partial change: 50% of the weights changed (attention, lm_head, mlp), in 77 of 80 layers.
Tensors changed: Embeddings 0/1, Norms 0/161, MLP 120/240, Attention 160/320, lm_head 1/1
11 derivatives checked
- bartowski/EVA-LLaMA-3.33-70B-v0.0-GGUF: Quantization
- mradermacher/EVA-LLaMA-3.33-70B-v0.0-i1-GGUF: Quantization
- Mawdistical/Anthrobomination-70B: Full fine-tune
- mradermacher/EVA-LLaMA-3.33-70B-v0.0-GGUF: Quantization
- Steelskull/L3.3-Electra-R1-70b: Full fine-tune
- EVA-UNIT-01/EVA-LLaMA-3.33-70B-v0.1: Full fine-tune
- SteelStorage/L3.3-MS-Evayale-70B: Partial fine-tune
- sophosympatheia/Sophos-eva-euryale-v1.0: Full fine-tune
- nitky/Llama-3.3-SuperSwallowX-70B-RP-v0.1: Full fine-tune
- TheSkullery/Unnamed-Exp-70b-v0.6: Full fine-tune
- ArtusDev/L3.3-Electra-R1-70b-Elarablation: Full fine-tune
1 other repos hold byte-for-byte identical weight files.