{ "architectures": [ "Qwen3ForCausalLM" ], "attention_bias": false, "attention_dropout": 0.0, "bos_token_id": 151643, "dtype": "bfloat16", "eos_token_id": 151645, "head_dim": 128, "hidden_act": "silu", "hidden_size": 2560, "initializer_range": 0.02, "intermediate_size": 9728, "layer_types": [ "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention", "full_attention" ], "max_position_embeddings": 40960, "max_window_layers": 36, "model_type": "qwen3", "num_attention_heads": 32, "num_hidden_layers": 36, "num_key_value_heads": 8, "quantization_config": { "config_groups": { "group_0": { "format": "pack-quantized", "input_activations": { "actorder": null, "block_structure": null, "dynamic": true, "group_size": null, "num_bits": 8, "observer": null, "observer_kwargs": {}, "scale_dtype": null, "strategy": "token", "symmetric": true, "type": "int", "zp_dtype": null }, "output_activations": null, "targets": [ "re:model\\.layers\\.0\\..*_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 7, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": null, "strategy": "channel", "symmetric": true, "type": "int", "zp_dtype": null } }, "group_1": { "format": "pack-quantized", "input_activations": null, "output_activations": null, "targets": [ "re:model\\.layers\\.1\\..*_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 4, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": null, "strategy": "channel", "symmetric": true, "type": "int", "zp_dtype": null } }, "group_10": { "format": "pack-quantized", "input_activations": null, "output_activations": null, "targets": [ "re:model\\.layers\\.10\\..*_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 4, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": null, "strategy": "channel", "symmetric": true, "type": "int", "zp_dtype": null } }, "group_11": { "format": "pack-quantized", "input_activations": null, "output_activations": null, "targets": [ "re:model\\.layers\\.11\\..*_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 5, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": null, "strategy": "channel", "symmetric": true, "type": "int", "zp_dtype": null } }, "group_12": { "format": "pack-quantized", "input_activations": null, "output_activations": null, "targets": [ "re:model\\.layers\\.12\\..*_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 6, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": null, "strategy": "channel", "symmetric": true, "type": "int", "zp_dtype": null } }, "group_13": { "format": "pack-quantized", "input_activations": null, "output_activations": null, "targets": [ "re:model\\.layers\\.13\\..*_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 7, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": null, "strategy": "channel", "symmetric": true, "type": "int", "zp_dtype": null } }, "group_14": { "format": "pack-quantized", "input_activations": null, "output_activations": null, "targets": [ "re:model\\.layers\\.14\\..*_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 8, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": null, "strategy": "channel", "symmetric": true, "type": "int", "zp_dtype": null } }, "group_15": { "format": "pack-quantized", "input_activations": { "actorder": null, "block_structure": null, "dynamic": true, "group_size": null, "num_bits": 8, "observer": null, "observer_kwargs": {}, "scale_dtype": null, "strategy": "token", "symmetric": true, "type": "int", "zp_dtype": null }, "output_activations": null, "targets": [ "re:model\\.layers\\.15\\..*_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 4, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": null, "strategy": "channel", "symmetric": true, "type": "int", "zp_dtype": null } }, "group_16": { "format": "pack-quantized", "input_activations": { "actorder": null, "block_structure": null, "dynamic": true, "group_size": null, "num_bits": 8, "observer": null, "observer_kwargs": {}, "scale_dtype": null, "strategy": "token", "symmetric": true, "type": "int", "zp_dtype": null }, "output_activations": null, "targets": [ "re:model\\.layers\\.16\\..*_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 5, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": null, "strategy": "channel", "symmetric": true, "type": "int", "zp_dtype": null } }, "group_17": { "format": "pack-quantized", "input_activations": { "actorder": null, "block_structure": null, "dynamic": true, "group_size": null, "num_bits": 8, "observer": null, "observer_kwargs": {}, "scale_dtype": null, "strategy": "token", "symmetric": true, "type": "int", "zp_dtype": null }, "output_activations": null, "targets": [ "re:model\\.layers\\.17\\..*_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 6, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": null, "strategy": "channel", "symmetric": true, "type": "int", "zp_dtype": null } }, "group_18": { "format": "pack-quantized", "input_activations": { "actorder": null, "block_structure": null, "dynamic": true, "group_size": null, "num_bits": 8, "observer": null, "observer_kwargs": {}, "scale_dtype": null, "strategy": "token", "symmetric": true, "type": "int", "zp_dtype": null }, "output_activations": null, "targets": [ "re:model\\.layers\\.18\\..*_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 7, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": null, "strategy": "channel", "symmetric": true, "type": "int", "zp_dtype": null } }, "group_19": { "format": "pack-quantized", "input_activations": null, "output_activations": null, "targets": [ "re:model\\.layers\\.19\\..*_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 4, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": null, "strategy": "channel", "symmetric": true, "type": "int", "zp_dtype": null } }, "group_2": { "format": "pack-quantized", "input_activations": null, "output_activations": null, "targets": [ "re:model\\.layers\\.2\\..*_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 5, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": null, "strategy": "channel", "symmetric": true, "type": "int", "zp_dtype": null } }, "group_20": { "format": "pack-quantized", "input_activations": null, "output_activations": null, "targets": [ "re:model\\.layers\\.20\\..*_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 5, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": null, "strategy": "channel", "symmetric": true, "type": "int", "zp_dtype": null } }, "group_21": { "format": "pack-quantized", "input_activations": null, "output_activations": null, "targets": [ "re:model\\.layers\\.21\\..*_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 6, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": null, "strategy": "channel", "symmetric": true, "type": "int", "zp_dtype": null } }, "group_22": { "format": "pack-quantized", "input_activations": null, "output_activations": null, "targets": [ "re:model\\.layers\\.22\\..*_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 7, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": null, "strategy": "channel", "symmetric": true, "type": "int", "zp_dtype": null } }, "group_23": { "format": "pack-quantized", "input_activations": null, "output_activations": null, "targets": [ "re:model\\.layers\\.23\\..*_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 8, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": null, "strategy": "channel", "symmetric": true, "type": "int", "zp_dtype": null } }, "group_24": { "format": "pack-quantized", "input_activations": { "actorder": null, "block_structure": null, "dynamic": true, "group_size": null, "num_bits": 8, "observer": null, "observer_kwargs": {}, "scale_dtype": null, "strategy": "token", "symmetric": true, "type": "int", "zp_dtype": null }, "output_activations": null, "targets": [ "re:model\\.layers\\.24\\..*_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 4, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": null, "strategy": "channel", "symmetric": true, "type": "int", "zp_dtype": null } }, "group_25": { "format": "pack-quantized", "input_activations": { "actorder": null, "block_structure": null, "dynamic": true, "group_size": null, "num_bits": 8, "observer": null, "observer_kwargs": {}, "scale_dtype": null, "strategy": "token", "symmetric": true, "type": "int", "zp_dtype": null }, "output_activations": null, "targets": [ "re:model\\.layers\\.25\\..*_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 5, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": null, "strategy": "channel", "symmetric": true, "type": "int", "zp_dtype": null } }, "group_26": { "format": "pack-quantized", "input_activations": { "actorder": null, "block_structure": null, "dynamic": true, "group_size": null, "num_bits": 8, "observer": null, "observer_kwargs": {}, "scale_dtype": null, "strategy": "token", "symmetric": true, "type": "int", "zp_dtype": null }, "output_activations": null, "targets": [ "re:model\\.layers\\.26\\..*_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 6, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": null, "strategy": "channel", "symmetric": true, "type": "int", "zp_dtype": null } }, "group_27": { "format": "pack-quantized", "input_activations": { "actorder": null, "block_structure": null, "dynamic": true, "group_size": null, "num_bits": 8, "observer": null, "observer_kwargs": {}, "scale_dtype": null, "strategy": "token", "symmetric": true, "type": "int", "zp_dtype": null }, "output_activations": null, "targets": [ "re:model\\.layers\\.27\\..*_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 7, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": null, "strategy": "channel", "symmetric": true, "type": "int", "zp_dtype": null } }, "group_28": { "format": "pack-quantized", "input_activations": null, "output_activations": null, "targets": [ "re:model\\.layers\\.28\\..*_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 4, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": null, "strategy": "channel", "symmetric": true, "type": "int", "zp_dtype": null } }, "group_29": { "format": "pack-quantized", "input_activations": null, "output_activations": null, "targets": [ "re:model\\.layers\\.29\\..*_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 5, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": null, "strategy": "channel", "symmetric": true, "type": "int", "zp_dtype": null } }, "group_3": { "format": "pack-quantized", "input_activations": null, "output_activations": null, "targets": [ "re:model\\.layers\\.3\\..*_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 6, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": null, "strategy": "channel", "symmetric": true, "type": "int", "zp_dtype": null } }, "group_30": { "format": "pack-quantized", "input_activations": null, "output_activations": null, "targets": [ "re:model\\.layers\\.30\\..*_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 6, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": null, "strategy": "channel", "symmetric": true, "type": "int", "zp_dtype": null } }, "group_31": { "format": "pack-quantized", "input_activations": null, "output_activations": null, "targets": [ "re:model\\.layers\\.31\\..*_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 7, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": null, "strategy": "channel", "symmetric": true, "type": "int", "zp_dtype": null } }, "group_32": { "format": "pack-quantized", "input_activations": null, "output_activations": null, "targets": [ "re:model\\.layers\\.32\\..*_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 8, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": null, "strategy": "channel", "symmetric": true, "type": "int", "zp_dtype": null } }, "group_33": { "format": "pack-quantized", "input_activations": { "actorder": null, "block_structure": null, "dynamic": true, "group_size": null, "num_bits": 8, "observer": null, "observer_kwargs": {}, "scale_dtype": null, "strategy": "token", "symmetric": true, "type": "int", "zp_dtype": null }, "output_activations": null, "targets": [ "re:model\\.layers\\.33\\..*_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 4, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": null, "strategy": "channel", "symmetric": true, "type": "int", "zp_dtype": null } }, "group_34": { "format": "pack-quantized", "input_activations": { "actorder": null, "block_structure": null, "dynamic": true, "group_size": null, "num_bits": 8, "observer": null, "observer_kwargs": {}, "scale_dtype": null, "strategy": "token", "symmetric": true, "type": "int", "zp_dtype": null }, "output_activations": null, "targets": [ "re:model\\.layers\\.34\\..*_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 5, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": null, "strategy": "channel", "symmetric": true, "type": "int", "zp_dtype": null } }, "group_35": { "format": "pack-quantized", "input_activations": { "actorder": null, "block_structure": null, "dynamic": true, "group_size": null, "num_bits": 8, "observer": null, "observer_kwargs": {}, "scale_dtype": null, "strategy": "token", "symmetric": true, "type": "int", "zp_dtype": null }, "output_activations": null, "targets": [ "re:model\\.layers\\.35\\..*_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 6, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": null, "strategy": "channel", "symmetric": true, "type": "int", "zp_dtype": null } }, "group_4": { "format": "pack-quantized", "input_activations": null, "output_activations": null, "targets": [ "re:model\\.layers\\.4\\..*_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 7, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": null, "strategy": "channel", "symmetric": true, "type": "int", "zp_dtype": null } }, "group_5": { "format": "pack-quantized", "input_activations": null, "output_activations": null, "targets": [ "re:model\\.layers\\.5\\..*_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 8, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": null, "strategy": "channel", "symmetric": true, "type": "int", "zp_dtype": null } }, "group_6": { "format": "pack-quantized", "input_activations": { "actorder": null, "block_structure": null, "dynamic": true, "group_size": null, "num_bits": 8, "observer": null, "observer_kwargs": {}, "scale_dtype": null, "strategy": "token", "symmetric": true, "type": "int", "zp_dtype": null }, "output_activations": null, "targets": [ "re:model\\.layers\\.6\\..*_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 4, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": null, "strategy": "channel", "symmetric": true, "type": "int", "zp_dtype": null } }, "group_7": { "format": "pack-quantized", "input_activations": { "actorder": null, "block_structure": null, "dynamic": true, "group_size": null, "num_bits": 8, "observer": null, "observer_kwargs": {}, "scale_dtype": null, "strategy": "token", "symmetric": true, "type": "int", "zp_dtype": null }, "output_activations": null, "targets": [ "re:model\\.layers\\.7\\..*_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 5, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": null, "strategy": "channel", "symmetric": true, "type": "int", "zp_dtype": null } }, "group_8": { "format": "pack-quantized", "input_activations": { "actorder": null, "block_structure": null, "dynamic": true, "group_size": null, "num_bits": 8, "observer": null, "observer_kwargs": {}, "scale_dtype": null, "strategy": "token", "symmetric": true, "type": "int", "zp_dtype": null }, "output_activations": null, "targets": [ "re:model\\.layers\\.8\\..*_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 6, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": null, "strategy": "channel", "symmetric": true, "type": "int", "zp_dtype": null } }, "group_9": { "format": "pack-quantized", "input_activations": { "actorder": null, "block_structure": null, "dynamic": true, "group_size": null, "num_bits": 8, "observer": null, "observer_kwargs": {}, "scale_dtype": null, "strategy": "token", "symmetric": true, "type": "int", "zp_dtype": null }, "output_activations": null, "targets": [ "re:model\\.layers\\.9\\..*_proj$" ], "weights": { "actorder": null, "block_structure": null, "dynamic": false, "group_size": null, "num_bits": 7, "observer": "memoryless_minmax", "observer_kwargs": {}, "scale_dtype": null, "strategy": "channel", "symmetric": true, "type": "int", "zp_dtype": null } } }, "format": "pack-quantized", "global_compression_ratio": null, "ignore": [ "lm_head" ], "kv_cache_scheme": null, "quant_method": "compressed-tensors", "quantization_status": "compressed", "sparsity_config": {}, "transform_config": {}, "version": "0.17.2.dev12+g3a9d460" }, "rms_norm_eps": 1e-06, "rope_scaling": null, "rope_theta": 1000000, "sliding_window": null, "tie_word_embeddings": true, "transformers_version": "4.57.6", "use_cache": true, "use_sliding_window": false, "vocab_size": 151936 }