| { | |
| "_name_or_path": "deepseek-ai/DeepSeek-R1-Distill-Qwen-32B", | |
| "architectures": [ | |
| "Qwen2ForCausalLM" | |
| ], | |
| "attention_dropout": 0.0, | |
| "bos_token_id": 151643, | |
| "eos_token_id": 151643, | |
| "hidden_act": "silu", | |
| "hidden_size": 5120, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 27648, | |
| "max_position_embeddings": 131072, | |
| "max_window_layers": 64, | |
| "model_type": "qwen2", | |
| "num_attention_heads": 40, | |
| "num_hidden_layers": 64, | |
| "num_key_value_heads": 8, | |
| "pad_token_id": 151654, | |
| "quantization_config": { | |
| "_load_in_4bit": true, | |
| "_load_in_8bit": false, | |
| "bnb_4bit_compute_dtype": "bfloat16", | |
| "bnb_4bit_quant_storage": "uint8", | |
| "bnb_4bit_quant_type": "nf4", | |
| "bnb_4bit_use_double_quant": true, | |
| "llm_int8_enable_fp32_cpu_offload": false, | |
| "llm_int8_has_fp16_weight": false, | |
| "llm_int8_skip_modules": [ | |
| "lm_head", | |
| "multi_modal_projector", | |
| "merger", | |
| "modality_projection", | |
| "model.layers.0.self_attn", | |
| "model.layers.0.mlp", | |
| "model.layers.1.mlp", | |
| "model.layers.2.self_attn", | |
| "model.layers.3.self_attn", | |
| "model.layers.3.mlp", | |
| "model.layers.4.self_attn", | |
| "model.layers.4.mlp", | |
| "model.layers.5.mlp", | |
| "model.layers.6.mlp", | |
| "model.layers.38.mlp", | |
| "model.layers.39.mlp", | |
| "model.layers.43.mlp", | |
| "model.layers.47.mlp", | |
| "model.layers.48.mlp", | |
| "model.layers.49.mlp", | |
| "model.layers.50.mlp", | |
| "model.layers.51.mlp", | |
| "model.layers.52.mlp", | |
| "model.layers.53.mlp", | |
| "model.layers.54.mlp", | |
| "model.layers.55.mlp", | |
| "model.layers.56.mlp", | |
| "model.layers.57.mlp", | |
| "model.layers.58.mlp", | |
| "model.layers.59.mlp", | |
| "model.layers.60.mlp", | |
| "model.layers.61.mlp", | |
| "model.layers.62.mlp", | |
| "model.layers.39.self_attn.o_proj", | |
| "model.layers.40.self_attn.o_proj", | |
| "model.layers.43.self_attn.o_proj" | |
| ], | |
| "llm_int8_threshold": 6.0, | |
| "load_in_4bit": true, | |
| "load_in_8bit": false, | |
| "quant_method": "bitsandbytes" | |
| }, | |
| "rms_norm_eps": 1e-05, | |
| "rope_scaling": null, | |
| "rope_theta": 1000000.0, | |
| "sliding_window": null, | |
| "tie_word_embeddings": false, | |
| "torch_dtype": "bfloat16", | |
| "transformers_version": "4.48.1", | |
| "unsloth_fixed": true, | |
| "use_cache": true, | |
| "use_sliding_window": false, | |
| "vocab_size": 152064 | |
| } | |