hotdogs/Qwen35B-Agent-R2
LoRA-style vs Qwen/Qwen-AgentWorld-35B-A3B. 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 40 of 40 layers; norms, biases and embeddings are untouched.
Tensors changed: MLP 120/280, Attention 40/280, Norms 0/131, Embeddings 0/1, lm_head 0/1
4 derivatives checked
- mradermacher/Qwen35B-Agent-R2-i1-GGUF: Quantization
- mradermacher/Qwen35B-Agent-R2-GGUF: Quantization
- hotdogs/Qwen35b-agent-R2O3: Partial fine-tune
- hotdogs/Qwen35-Agent-R2A103: Full fine-tune