codefuse-ai/F2LLM-v2-0.6B-Preview
Full fine-tune vs Qwen/Qwen3-0.6B. Fine-tune: 100% of the weights changed, across 28 of 28 layers.
Tensors changed: Embeddings 1/1, Norms 106/113, MLP 84/84, Attention 112/112
5 derivatives checked
- codefuse-ai/F2LLM-v2-0.6B: Full fine-tune
- codefuse-ai/ML-Embed-0.6B: Full fine-tune
- codefuse-ai/F2LLM-v2-0.6B-Preview-Pruned-330M: Different architecture
- codefuse-ai/F2LLM-v2-0.6B-Preview-Pruned-160M: Different architecture
- codefuse-ai/F2LLM-v2-0.6B-Preview-Pruned-80M: Different architecture