meta-llama/Llama-3.3-70B-Instruct
Full fine-tune vs meta-llama/Llama-3.1-70B. Fine-tune: 100% of the weights changed, across 80 of 80 layers.
Tensors changed: Embeddings 1/1, Norms 161/161, MLP 240/240, Attention 320/320, lm_head 1/1
153 derivatives checked
- MaziyarPanahi/Llama-3.3-70B-Instruct-GGUF: Same weights, converted
- RedHatAI/Llama-3.3-70B-Instruct-FP8-dynamic: Quantization
- cortecs/Llama-3.3-70B-Instruct-FP8-Dynamic: Quantization
- unsloth/Llama-3.3-70B-Instruct: Identical copy
- ibnzterrell/Meta-Llama-3.3-70B-Instruct-AWQ-INT4: Quantization
- unsloth/Llama-3.3-70B-Instruct-GGUF: Same weights, converted
- bartowski/Llama-3.3-70B-Instruct-GGUF: Same weights, converted
- nvidia/Llama-3.3-70B-Instruct-FP8: Quantization
- amd/Llama-3.3-70B-Instruct-FP8-KV: Quantization
- nvidia/Llama-3.3-70B-Instruct-NVFP4: Quantization
- huihui-ai/Llama-3.3-70B-Instruct-abliterated: Abliterated
- lmstudio-community/Llama-3.3-70B-Instruct-GGUF: Quantization
- unsloth/Llama-3.3-70B-Instruct-bnb-4bit: Quantization
- RedHatAI/Llama-3.3-70B-Instruct-quantized.w4a16: Quantization
- auditing-agents/llama_70b_transcripts_only_then_redteam_high_ai_welfare_poisoning: Adapter
- mlx-community/Llama-3.3-70B-Instruct-4bit: Quantization
- pearl-ai/Llama-3.3-70B-Instruct-pearl: Quantization
- kosbu/Llama-3.3-70B-Instruct-AWQ: Quantization
- mlx-community/Llama-3.3-70B-Instruct-8bit: Quantization
- amd/Llama-3.3-70B-Instruct-MXFP4-Preview: Quantization
- TheDrummer/Anubis-70B-v1.2-GGUF: Full fine-tune
- RedHatAI/Llama-3.3-70B-Instruct: Identical copy
- Trendyol/Trendyol-Cybersecurity-LLM-v2-70B-Q4_K_M: Quantization
- Thaurock/Llama-3.3-70B-Instruct-abliterated-GGUF: Abliterated
- mlx-community/llama-3.3-70b-instruct-fp16: Same weights, converted
- AtlaAI/Selene-1-Llama-3.3-70B: Full fine-tune
- axana-labs/Llama-3.3-70B-Instruct-bnb-4bit: Quantization
- backyardai/Llama-3.3-70B-Instruct-GGUF: Same weights, converted
- Sao10K/L3.3-70B-Euryale-v2.3: LoRA-style
- mradermacher/Llama-3.3-70B-Instruct-i1-GGUF: Quantization
17 other repos hold byte-for-byte identical weight files.