Upload folder using huggingface_hub

_config.json

@@ -0,0 +1,410 @@
+{
+  "architectures": [
+    "KimiLinearForCausalLM"
+  ],
+  "auto_map": {
+    "AutoConfig": "configuration_kimi.KimiLinearConfig",
+    "AutoModel": "modeling_kimi.KimiLinearModel",
+    "AutoModelForCausalLM": "modeling_kimi.KimiLinearForCausalLM"
+  },
+  "bos_token_id": 163584,
+  "dtype": "bfloat16",
+  "eos_token_id": 163586,
+  "first_k_dense_replace": 1,
+  "head_dim": 72,
+  "hidden_act": "silu",
+  "hidden_size": 2304,
+  "initializer_range": 0.02,
+  "intermediate_size": 9216,
+  "kv_lora_rank": 512,
+  "linear_attn_config": {
+    "full_attn_layers": [
+      4,
+      8,
+      12,
+      16,
+      20,
+      24,
+      27
+    ],
+    "head_dim": 128,
+    "kda_layers": [
+      1,
+      2,
+      3,
+      5,
+      6,
+      7,
+      9,
+      10,
+      11,
+      13,
+      14,
+      15,
+      17,
+      18,
+      19,
+      21,
+      22,
+      23,
+      25,
+      26
+    ],
+    "num_heads": 32,
+    "short_conv_kernel_size": 4
+  },
+  "mla_use_nope": true,
+  "model_max_length": 1048576,
+  "model_type": "kimi_linear",
+  "moe_intermediate_size": 1024,
+  "moe_layer_freq": 1,
+  "moe_renormalize": true,
+  "moe_router_activation_func": "sigmoid",
+  "num_attention_heads": 32,
+  "num_expert_group": 1,
+  "num_experts": 256,
+  "num_experts_per_token": 8,
+  "num_hidden_layers": 27,
+  "num_key_value_heads": 32,
+  "num_nextn_predict_layers": 0,
+  "num_shared_experts": 1,
+  "pad_token_id": 163839,
+  "q_lora_rank": null,
+  "qk_nope_head_dim": 128,
+  "qk_rope_head_dim": 64,
+  "quantization_config": {
+    "config_groups": {
+      "group_0": {
+        "format": "pack-quantized",
+        "input_activations": null,
+        "output_activations": null,
+        "targets": [
+          "Linear"
+        ],
+        "weights": {
+          "actorder": null,
+          "block_structure": null,
+          "dynamic": false,
+          "group_size": 32,
+          "num_bits": 4,
+          "observer": "mse",
+          "observer_kwargs": {},
+          "strategy": "group",
+          "symmetric": true,
+          "type": "int"
+        }
+      }
+    },
+    "format": "pack-quantized",
+    "global_compression_ratio": null,
+    "ignore": [
+      "model.layers.0.self_attn.q_proj",
+      "model.layers.0.self_attn.k_proj",
+      "model.layers.0.self_attn.v_proj",
+      "model.layers.0.self_attn.f_a_proj",
+      "model.layers.0.self_attn.f_b_proj",
+      "model.layers.0.self_attn.b_proj",
+      "model.layers.0.self_attn.g_a_proj",
+      "model.layers.0.self_attn.g_b_proj",
+      "model.layers.0.self_attn.o_proj",
+      "model.layers.0.mlp.gate_proj",
+      "model.layers.0.mlp.up_proj",
+      "model.layers.0.mlp.down_proj",
+      "model.layers.1.self_attn.q_proj",
+      "model.layers.1.self_attn.k_proj",
+      "model.layers.1.self_attn.v_proj",
+      "model.layers.1.self_attn.f_a_proj",
+      "model.layers.1.self_attn.f_b_proj",
+      "model.layers.1.self_attn.b_proj",
+      "model.layers.1.self_attn.g_a_proj",
+      "model.layers.1.self_attn.g_b_proj",
+      "model.layers.1.self_attn.o_proj",
+      "model.layers.1.block_sparse_moe.shared_experts.gate_proj",
+      "model.layers.1.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.1.block_sparse_moe.shared_experts.down_proj",
+      "model.layers.2.self_attn.q_proj",
+      "model.layers.2.self_attn.k_proj",
+      "model.layers.2.self_attn.v_proj",
+      "model.layers.2.self_attn.f_a_proj",
+      "model.layers.2.self_attn.f_b_proj",
+      "model.layers.2.self_attn.b_proj",
+      "model.layers.2.self_attn.g_a_proj",
+      "model.layers.2.self_attn.g_b_proj",
+      "model.layers.2.self_attn.o_proj",
+      "model.layers.2.block_sparse_moe.shared_experts.gate_proj",
+      "model.layers.2.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.2.block_sparse_moe.shared_experts.down_proj",
+      "model.layers.3.self_attn.q_proj",
+      "model.layers.3.self_attn.kv_a_proj_with_mqa",
+      "model.layers.3.self_attn.kv_b_proj",
+      "model.layers.3.self_attn.o_proj",
+      "model.layers.3.block_sparse_moe.shared_experts.gate_proj",
+      "model.layers.3.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.3.block_sparse_moe.shared_experts.down_proj",
+      "model.layers.4.self_attn.q_proj",
+      "model.layers.4.self_attn.k_proj",
+      "model.layers.4.self_attn.v_proj",
+      "model.layers.4.self_attn.f_a_proj",
+      "model.layers.4.self_attn.f_b_proj",
+      "model.layers.4.self_attn.b_proj",
+      "model.layers.4.self_attn.g_a_proj",
+      "model.layers.4.self_attn.g_b_proj",
+      "model.layers.4.self_attn.o_proj",
+      "model.layers.4.block_sparse_moe.shared_experts.gate_proj",
+      "model.layers.4.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.4.block_sparse_moe.shared_experts.down_proj",
+      "model.layers.5.self_attn.q_proj",
+      "model.layers.5.self_attn.k_proj",
+      "model.layers.5.self_attn.v_proj",
+      "model.layers.5.self_attn.f_a_proj",
+      "model.layers.5.self_attn.f_b_proj",
+      "model.layers.5.self_attn.b_proj",
+      "model.layers.5.self_attn.g_a_proj",
+      "model.layers.5.self_attn.g_b_proj",
+      "model.layers.5.self_attn.o_proj",
+      "model.layers.5.block_sparse_moe.shared_experts.gate_proj",
+      "model.layers.5.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.5.block_sparse_moe.shared_experts.down_proj",
+      "model.layers.6.self_attn.q_proj",
+      "model.layers.6.self_attn.k_proj",
+      "model.layers.6.self_attn.v_proj",
+      "model.layers.6.self_attn.f_a_proj",
+      "model.layers.6.self_attn.f_b_proj",
+      "model.layers.6.self_attn.b_proj",
+      "model.layers.6.self_attn.g_a_proj",
+      "model.layers.6.self_attn.g_b_proj",
+      "model.layers.6.self_attn.o_proj",
+      "model.layers.6.block_sparse_moe.shared_experts.gate_proj",
+      "model.layers.6.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.6.block_sparse_moe.shared_experts.down_proj",
+      "model.layers.7.self_attn.q_proj",
+      "model.layers.7.self_attn.kv_a_proj_with_mqa",
+      "model.layers.7.self_attn.kv_b_proj",
+      "model.layers.7.self_attn.o_proj",
+      "model.layers.7.block_sparse_moe.shared_experts.gate_proj",
+      "model.layers.7.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.7.block_sparse_moe.shared_experts.down_proj",
+      "model.layers.8.self_attn.q_proj",
+      "model.layers.8.self_attn.k_proj",
+      "model.layers.8.self_attn.v_proj",
+      "model.layers.8.self_attn.f_a_proj",
+      "model.layers.8.self_attn.f_b_proj",
+      "model.layers.8.self_attn.b_proj",
+      "model.layers.8.self_attn.g_a_proj",
+      "model.layers.8.self_attn.g_b_proj",
+      "model.layers.8.self_attn.o_proj",
+      "model.layers.8.block_sparse_moe.shared_experts.gate_proj",
+      "model.layers.8.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.8.block_sparse_moe.shared_experts.down_proj",
+      "model.layers.9.self_attn.q_proj",
+      "model.layers.9.self_attn.k_proj",
+      "model.layers.9.self_attn.v_proj",
+      "model.layers.9.self_attn.f_a_proj",
+      "model.layers.9.self_attn.f_b_proj",
+      "model.layers.9.self_attn.b_proj",
+      "model.layers.9.self_attn.g_a_proj",
+      "model.layers.9.self_attn.g_b_proj",
+      "model.layers.9.self_attn.o_proj",
+      "model.layers.9.block_sparse_moe.shared_experts.gate_proj",
+      "model.layers.9.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.9.block_sparse_moe.shared_experts.down_proj",
+      "model.layers.10.self_attn.q_proj",
+      "model.layers.10.self_attn.k_proj",
+      "model.layers.10.self_attn.v_proj",
+      "model.layers.10.self_attn.f_a_proj",
+      "model.layers.10.self_attn.f_b_proj",
+      "model.layers.10.self_attn.b_proj",
+      "model.layers.10.self_attn.g_a_proj",
+      "model.layers.10.self_attn.g_b_proj",
+      "model.layers.10.self_attn.o_proj",
+      "model.layers.10.block_sparse_moe.shared_experts.gate_proj",
+      "model.layers.10.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.10.block_sparse_moe.shared_experts.down_proj",
+      "model.layers.11.self_attn.q_proj",
+      "model.layers.11.self_attn.kv_a_proj_with_mqa",
+      "model.layers.11.self_attn.kv_b_proj",
+      "model.layers.11.self_attn.o_proj",
+      "model.layers.11.block_sparse_moe.shared_experts.gate_proj",
+      "model.layers.11.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.11.block_sparse_moe.shared_experts.down_proj",
+      "model.layers.12.self_attn.q_proj",
+      "model.layers.12.self_attn.k_proj",
+      "model.layers.12.self_attn.v_proj",
+      "model.layers.12.self_attn.f_a_proj",
+      "model.layers.12.self_attn.f_b_proj",
+      "model.layers.12.self_attn.b_proj",
+      "model.layers.12.self_attn.g_a_proj",
+      "model.layers.12.self_attn.g_b_proj",
+      "model.layers.12.self_attn.o_proj",
+      "model.layers.12.block_sparse_moe.shared_experts.gate_proj",
+      "model.layers.12.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.12.block_sparse_moe.shared_experts.down_proj",
+      "model.layers.13.self_attn.q_proj",
+      "model.layers.13.self_attn.k_proj",
+      "model.layers.13.self_attn.v_proj",
+      "model.layers.13.self_attn.f_a_proj",
+      "model.layers.13.self_attn.f_b_proj",
+      "model.layers.13.self_attn.b_proj",
+      "model.layers.13.self_attn.g_a_proj",
+      "model.layers.13.self_attn.g_b_proj",
+      "model.layers.13.self_attn.o_proj",
+      "model.layers.13.block_sparse_moe.shared_experts.gate_proj",
+      "model.layers.13.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.13.block_sparse_moe.shared_experts.down_proj",
+      "model.layers.14.self_attn.q_proj",
+      "model.layers.14.self_attn.k_proj",
+      "model.layers.14.self_attn.v_proj",
+      "model.layers.14.self_attn.f_a_proj",
+      "model.layers.14.self_attn.f_b_proj",
+      "model.layers.14.self_attn.b_proj",
+      "model.layers.14.self_attn.g_a_proj",
+      "model.layers.14.self_attn.g_b_proj",
+      "model.layers.14.self_attn.o_proj",
+      "model.layers.14.block_sparse_moe.shared_experts.gate_proj",
+      "model.layers.14.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.14.block_sparse_moe.shared_experts.down_proj",
+      "model.layers.15.self_attn.q_proj",
+      "model.layers.15.self_attn.kv_a_proj_with_mqa",
+      "model.layers.15.self_attn.kv_b_proj",
+      "model.layers.15.self_attn.o_proj",
+      "model.layers.15.block_sparse_moe.shared_experts.gate_proj",
+      "model.layers.15.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.15.block_sparse_moe.shared_experts.down_proj",
+      "model.layers.16.self_attn.q_proj",
+      "model.layers.16.self_attn.k_proj",
+      "model.layers.16.self_attn.v_proj",
+      "model.layers.16.self_attn.f_a_proj",
+      "model.layers.16.self_attn.f_b_proj",
+      "model.layers.16.self_attn.b_proj",
+      "model.layers.16.self_attn.g_a_proj",
+      "model.layers.16.self_attn.g_b_proj",
+      "model.layers.16.self_attn.o_proj",
+      "model.layers.16.block_sparse_moe.shared_experts.gate_proj",
+      "model.layers.16.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.16.block_sparse_moe.shared_experts.down_proj",
+      "model.layers.17.self_attn.q_proj",
+      "model.layers.17.self_attn.k_proj",
+      "model.layers.17.self_attn.v_proj",
+      "model.layers.17.self_attn.f_a_proj",
+      "model.layers.17.self_attn.f_b_proj",
+      "model.layers.17.self_attn.b_proj",
+      "model.layers.17.self_attn.g_a_proj",
+      "model.layers.17.self_attn.g_b_proj",
+      "model.layers.17.self_attn.o_proj",
+      "model.layers.17.block_sparse_moe.shared_experts.gate_proj",
+      "model.layers.17.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.17.block_sparse_moe.shared_experts.down_proj",
+      "model.layers.18.self_attn.q_proj",
+      "model.layers.18.self_attn.k_proj",
+      "model.layers.18.self_attn.v_proj",
+      "model.layers.18.self_attn.f_a_proj",
+      "model.layers.18.self_attn.f_b_proj",
+      "model.layers.18.self_attn.b_proj",
+      "model.layers.18.self_attn.g_a_proj",
+      "model.layers.18.self_attn.g_b_proj",
+      "model.layers.18.self_attn.o_proj",
+      "model.layers.18.block_sparse_moe.shared_experts.gate_proj",
+      "model.layers.18.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.18.block_sparse_moe.shared_experts.down_proj",
+      "model.layers.19.self_attn.q_proj",
+      "model.layers.19.self_attn.kv_a_proj_with_mqa",
+      "model.layers.19.self_attn.kv_b_proj",
+      "model.layers.19.self_attn.o_proj",
+      "model.layers.19.block_sparse_moe.shared_experts.gate_proj",
+      "model.layers.19.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.19.block_sparse_moe.shared_experts.down_proj",
+      "model.layers.20.self_attn.q_proj",
+      "model.layers.20.self_attn.k_proj",
+      "model.layers.20.self_attn.v_proj",
+      "model.layers.20.self_attn.f_a_proj",
+      "model.layers.20.self_attn.f_b_proj",
+      "model.layers.20.self_attn.b_proj",
+      "model.layers.20.self_attn.g_a_proj",
+      "model.layers.20.self_attn.g_b_proj",
+      "model.layers.20.self_attn.o_proj",
+      "model.layers.20.block_sparse_moe.shared_experts.gate_proj",
+      "model.layers.20.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.20.block_sparse_moe.shared_experts.down_proj",
+      "model.layers.21.self_attn.q_proj",
+      "model.layers.21.self_attn.k_proj",
+      "model.layers.21.self_attn.v_proj",
+      "model.layers.21.self_attn.f_a_proj",
+      "model.layers.21.self_attn.f_b_proj",
+      "model.layers.21.self_attn.b_proj",
+      "model.layers.21.self_attn.g_a_proj",
+      "model.layers.21.self_attn.g_b_proj",
+      "model.layers.21.self_attn.o_proj",
+      "model.layers.21.block_sparse_moe.shared_experts.gate_proj",
+      "model.layers.21.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.21.block_sparse_moe.shared_experts.down_proj",
+      "model.layers.22.self_attn.q_proj",
+      "model.layers.22.self_attn.k_proj",
+      "model.layers.22.self_attn.v_proj",
+      "model.layers.22.self_attn.f_a_proj",
+      "model.layers.22.self_attn.f_b_proj",
+      "model.layers.22.self_attn.b_proj",
+      "model.layers.22.self_attn.g_a_proj",
+      "model.layers.22.self_attn.g_b_proj",
+      "model.layers.22.self_attn.o_proj",
+      "model.layers.22.block_sparse_moe.shared_experts.gate_proj",
+      "model.layers.22.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.22.block_sparse_moe.shared_experts.down_proj",
+      "model.layers.23.self_attn.q_proj",
+      "model.layers.23.self_attn.kv_a_proj_with_mqa",
+      "model.layers.23.self_attn.kv_b_proj",
+      "model.layers.23.self_attn.o_proj",
+      "model.layers.23.block_sparse_moe.shared_experts.gate_proj",
+      "model.layers.23.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.23.block_sparse_moe.shared_experts.down_proj",
+      "model.layers.24.self_attn.q_proj",
+      "model.layers.24.self_attn.k_proj",
+      "model.layers.24.self_attn.v_proj",
+      "model.layers.24.self_attn.f_a_proj",
+      "model.layers.24.self_attn.f_b_proj",
+      "model.layers.24.self_attn.b_proj",
+      "model.layers.24.self_attn.g_a_proj",
+      "model.layers.24.self_attn.g_b_proj",
+      "model.layers.24.self_attn.o_proj",
+      "model.layers.24.block_sparse_moe.shared_experts.gate_proj",
+      "model.layers.24.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.24.block_sparse_moe.shared_experts.down_proj",
+      "model.layers.25.self_attn.q_proj",
+      "model.layers.25.self_attn.k_proj",
+      "model.layers.25.self_attn.v_proj",
+      "model.layers.25.self_attn.f_a_proj",
+      "model.layers.25.self_attn.f_b_proj",
+      "model.layers.25.self_attn.b_proj",
+      "model.layers.25.self_attn.g_a_proj",
+      "model.layers.25.self_attn.g_b_proj",
+      "model.layers.25.self_attn.o_proj",
+      "model.layers.25.block_sparse_moe.shared_experts.gate_proj",
+      "model.layers.25.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.25.block_sparse_moe.shared_experts.down_proj",
+      "model.layers.26.self_attn.q_proj",
+      "model.layers.26.self_attn.kv_a_proj_with_mqa",
+      "model.layers.26.self_attn.kv_b_proj",
+      "model.layers.26.self_attn.o_proj",
+      "model.layers.26.block_sparse_moe.shared_experts.gate_proj",
+      "model.layers.26.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.26.block_sparse_moe.shared_experts.down_proj",
+      "lm_head"
+    ],
+    "kv_cache_scheme": null,
+    "quant_method": "compressed-tensors",
+    "quantization_status": "compressed",
+    "sparsity_config": {},
+    "transform_config": {},
+    "version": "0.12.3.dev20+gd429903"
+  },
+  "rms_norm_eps": 1e-05,
+  "rope_scaling": null,
+  "rope_theta": 10000.0,
+  "routed_scaling_factor": 2.446,
+  "tie_word_embeddings": false,
+  "topk_group": 1,
+  "transformers_version": "4.57.0",
+  "use_cache": true,
+  "use_grouped_topk": true,
+  "v_head_dim": 128,
+  "vocab_size": 163840
+}

config.json

@@ -109,6 +109,7 @@
       "model.layers.0.self_attn.o_proj",
       "model.layers.0.mlp.gate_proj",
       "model.layers.0.mlp.up_proj",
+      "model.layers.0.mlp.gate_up_proj",
       "model.layers.0.mlp.down_proj",
       "model.layers.1.self_attn.q_proj",
       "model.layers.1.self_attn.k_proj",
@@ -121,6 +122,7 @@
       "model.layers.1.self_attn.o_proj",
       "model.layers.1.block_sparse_moe.shared_experts.gate_proj",
       "model.layers.1.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.1.block_sparse_moe.shared_experts.gate_up_proj",
       "model.layers.1.block_sparse_moe.shared_experts.down_proj",
       "model.layers.2.self_attn.q_proj",
       "model.layers.2.self_attn.k_proj",
@@ -133,6 +135,7 @@
       "model.layers.2.self_attn.o_proj",
       "model.layers.2.block_sparse_moe.shared_experts.gate_proj",
       "model.layers.2.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.2.block_sparse_moe.shared_experts.gate_up_proj",
       "model.layers.2.block_sparse_moe.shared_experts.down_proj",
       "model.layers.3.self_attn.q_proj",
       "model.layers.3.self_attn.kv_a_proj_with_mqa",
@@ -140,6 +143,7 @@
       "model.layers.3.self_attn.o_proj",
       "model.layers.3.block_sparse_moe.shared_experts.gate_proj",
       "model.layers.3.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.3.block_sparse_moe.shared_experts.gate_up_proj",
       "model.layers.3.block_sparse_moe.shared_experts.down_proj",
       "model.layers.4.self_attn.q_proj",
       "model.layers.4.self_attn.k_proj",
@@ -152,6 +156,7 @@
       "model.layers.4.self_attn.o_proj",
       "model.layers.4.block_sparse_moe.shared_experts.gate_proj",
       "model.layers.4.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.4.block_sparse_moe.shared_experts.gate_up_proj",
       "model.layers.4.block_sparse_moe.shared_experts.down_proj",
       "model.layers.5.self_attn.q_proj",
       "model.layers.5.self_attn.k_proj",
@@ -164,6 +169,7 @@
       "model.layers.5.self_attn.o_proj",
       "model.layers.5.block_sparse_moe.shared_experts.gate_proj",
       "model.layers.5.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.5.block_sparse_moe.shared_experts.gate_up_proj",
       "model.layers.5.block_sparse_moe.shared_experts.down_proj",
       "model.layers.6.self_attn.q_proj",
       "model.layers.6.self_attn.k_proj",
@@ -176,6 +182,7 @@
       "model.layers.6.self_attn.o_proj",
       "model.layers.6.block_sparse_moe.shared_experts.gate_proj",
       "model.layers.6.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.6.block_sparse_moe.shared_experts.gate_up_proj",
       "model.layers.6.block_sparse_moe.shared_experts.down_proj",
       "model.layers.7.self_attn.q_proj",
       "model.layers.7.self_attn.kv_a_proj_with_mqa",
@@ -183,6 +190,7 @@
       "model.layers.7.self_attn.o_proj",
       "model.layers.7.block_sparse_moe.shared_experts.gate_proj",
       "model.layers.7.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.7.block_sparse_moe.shared_experts.gate_up_proj",
       "model.layers.7.block_sparse_moe.shared_experts.down_proj",
       "model.layers.8.self_attn.q_proj",
       "model.layers.8.self_attn.k_proj",
@@ -195,6 +203,7 @@
       "model.layers.8.self_attn.o_proj",
       "model.layers.8.block_sparse_moe.shared_experts.gate_proj",
       "model.layers.8.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.8.block_sparse_moe.shared_experts.gate_up_proj",
       "model.layers.8.block_sparse_moe.shared_experts.down_proj",
       "model.layers.9.self_attn.q_proj",
       "model.layers.9.self_attn.k_proj",
@@ -207,6 +216,7 @@
       "model.layers.9.self_attn.o_proj",
       "model.layers.9.block_sparse_moe.shared_experts.gate_proj",
       "model.layers.9.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.9.block_sparse_moe.shared_experts.gate_up_proj",
       "model.layers.9.block_sparse_moe.shared_experts.down_proj",
       "model.layers.10.self_attn.q_proj",
       "model.layers.10.self_attn.k_proj",
@@ -219,6 +229,7 @@
       "model.layers.10.self_attn.o_proj",
       "model.layers.10.block_sparse_moe.shared_experts.gate_proj",
       "model.layers.10.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.10.block_sparse_moe.shared_experts.gate_up_proj",
       "model.layers.10.block_sparse_moe.shared_experts.down_proj",
       "model.layers.11.self_attn.q_proj",
       "model.layers.11.self_attn.kv_a_proj_with_mqa",
@@ -226,6 +237,7 @@
       "model.layers.11.self_attn.o_proj",
       "model.layers.11.block_sparse_moe.shared_experts.gate_proj",
       "model.layers.11.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.11.block_sparse_moe.shared_experts.gate_up_proj",
       "model.layers.11.block_sparse_moe.shared_experts.down_proj",
       "model.layers.12.self_attn.q_proj",
       "model.layers.12.self_attn.k_proj",
@@ -238,6 +250,7 @@
       "model.layers.12.self_attn.o_proj",
       "model.layers.12.block_sparse_moe.shared_experts.gate_proj",
       "model.layers.12.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.12.block_sparse_moe.shared_experts.gate_up_proj",
       "model.layers.12.block_sparse_moe.shared_experts.down_proj",
       "model.layers.13.self_attn.q_proj",
       "model.layers.13.self_attn.k_proj",
@@ -250,6 +263,7 @@
       "model.layers.13.self_attn.o_proj",
       "model.layers.13.block_sparse_moe.shared_experts.gate_proj",
       "model.layers.13.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.13.block_sparse_moe.shared_experts.gate_up_proj",
       "model.layers.13.block_sparse_moe.shared_experts.down_proj",
       "model.layers.14.self_attn.q_proj",
       "model.layers.14.self_attn.k_proj",
@@ -262,6 +276,7 @@
       "model.layers.14.self_attn.o_proj",
       "model.layers.14.block_sparse_moe.shared_experts.gate_proj",
       "model.layers.14.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.14.block_sparse_moe.shared_experts.gate_up_proj",
       "model.layers.14.block_sparse_moe.shared_experts.down_proj",
       "model.layers.15.self_attn.q_proj",
       "model.layers.15.self_attn.kv_a_proj_with_mqa",
@@ -269,6 +284,7 @@
       "model.layers.15.self_attn.o_proj",
       "model.layers.15.block_sparse_moe.shared_experts.gate_proj",
       "model.layers.15.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.15.block_sparse_moe.shared_experts.gate_up_proj",
       "model.layers.15.block_sparse_moe.shared_experts.down_proj",
       "model.layers.16.self_attn.q_proj",
       "model.layers.16.self_attn.k_proj",
@@ -281,6 +297,7 @@
       "model.layers.16.self_attn.o_proj",
       "model.layers.16.block_sparse_moe.shared_experts.gate_proj",
       "model.layers.16.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.16.block_sparse_moe.shared_experts.gate_up_proj",
       "model.layers.16.block_sparse_moe.shared_experts.down_proj",
       "model.layers.17.self_attn.q_proj",
       "model.layers.17.self_attn.k_proj",
@@ -293,6 +310,7 @@
       "model.layers.17.self_attn.o_proj",
       "model.layers.17.block_sparse_moe.shared_experts.gate_proj",
       "model.layers.17.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.17.block_sparse_moe.shared_experts.gate_up_proj",
       "model.layers.17.block_sparse_moe.shared_experts.down_proj",
       "model.layers.18.self_attn.q_proj",
       "model.layers.18.self_attn.k_proj",
@@ -305,6 +323,7 @@
       "model.layers.18.self_attn.o_proj",
       "model.layers.18.block_sparse_moe.shared_experts.gate_proj",
       "model.layers.18.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.18.block_sparse_moe.shared_experts.gate_up_proj",
       "model.layers.18.block_sparse_moe.shared_experts.down_proj",
       "model.layers.19.self_attn.q_proj",
       "model.layers.19.self_attn.kv_a_proj_with_mqa",
@@ -312,6 +331,7 @@
       "model.layers.19.self_attn.o_proj",
       "model.layers.19.block_sparse_moe.shared_experts.gate_proj",
       "model.layers.19.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.19.block_sparse_moe.shared_experts.gate_up_proj",
       "model.layers.19.block_sparse_moe.shared_experts.down_proj",
       "model.layers.20.self_attn.q_proj",
       "model.layers.20.self_attn.k_proj",
@@ -324,6 +344,7 @@
       "model.layers.20.self_attn.o_proj",
       "model.layers.20.block_sparse_moe.shared_experts.gate_proj",
       "model.layers.20.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.20.block_sparse_moe.shared_experts.gate_up_proj",
       "model.layers.20.block_sparse_moe.shared_experts.down_proj",
       "model.layers.21.self_attn.q_proj",
       "model.layers.21.self_attn.k_proj",
@@ -336,6 +357,7 @@
       "model.layers.21.self_attn.o_proj",
       "model.layers.21.block_sparse_moe.shared_experts.gate_proj",
       "model.layers.21.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.21.block_sparse_moe.shared_experts.gate_up_proj",
       "model.layers.21.block_sparse_moe.shared_experts.down_proj",
       "model.layers.22.self_attn.q_proj",
       "model.layers.22.self_attn.k_proj",
@@ -348,6 +370,7 @@
       "model.layers.22.self_attn.o_proj",
       "model.layers.22.block_sparse_moe.shared_experts.gate_proj",
       "model.layers.22.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.22.block_sparse_moe.shared_experts.gate_up_proj",
       "model.layers.22.block_sparse_moe.shared_experts.down_proj",
       "model.layers.23.self_attn.q_proj",
       "model.layers.23.self_attn.kv_a_proj_with_mqa",
@@ -355,6 +378,7 @@
       "model.layers.23.self_attn.o_proj",
       "model.layers.23.block_sparse_moe.shared_experts.gate_proj",
       "model.layers.23.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.23.block_sparse_moe.shared_experts.gate_up_proj",
       "model.layers.23.block_sparse_moe.shared_experts.down_proj",
       "model.layers.24.self_attn.q_proj",
       "model.layers.24.self_attn.k_proj",
@@ -367,6 +391,7 @@
       "model.layers.24.self_attn.o_proj",
       "model.layers.24.block_sparse_moe.shared_experts.gate_proj",
       "model.layers.24.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.24.block_sparse_moe.shared_experts.gate_up_proj",
       "model.layers.24.block_sparse_moe.shared_experts.down_proj",
       "model.layers.25.self_attn.q_proj",
       "model.layers.25.self_attn.k_proj",
@@ -379,6 +404,7 @@
       "model.layers.25.self_attn.o_proj",
       "model.layers.25.block_sparse_moe.shared_experts.gate_proj",
       "model.layers.25.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.25.block_sparse_moe.shared_experts.gate_up_proj",
       "model.layers.25.block_sparse_moe.shared_experts.down_proj",
       "model.layers.26.self_attn.q_proj",
       "model.layers.26.self_attn.kv_a_proj_with_mqa",
@@ -386,6 +412,7 @@
       "model.layers.26.self_attn.o_proj",
       "model.layers.26.block_sparse_moe.shared_experts.gate_proj",
       "model.layers.26.block_sparse_moe.shared_experts.up_proj",
+      "model.layers.26.block_sparse_moe.shared_experts.gate_up_proj",
       "model.layers.26.block_sparse_moe.shared_experts.down_proj",
       "lm_head"
     ],

modeling_kimi.py

@@ -0,0 +1,1028 @@
+import math
+from collections.abc import Callable
+from typing import Any, List, Optional, Tuple, Union
+
+import torch
+import torch.nn.functional as F
+import transformers
+from einops import rearrange
+from packaging import version
+from torch import nn
+from transformers.activations import ACT2FN
+from transformers.cache_utils import Cache
+from transformers.generation import GenerationMixin
+from transformers.masking_utils import create_causal_mask
+from transformers.modeling_flash_attention_utils import FlashAttentionKwargs
+from transformers.modeling_outputs import (BaseModelOutputWithPast,
+                                           CausalLMOutputWithPast)
+from transformers.modeling_utils import (ALL_ATTENTION_FUNCTIONS,
+                                         PreTrainedModel)
+from transformers.processing_utils import Unpack
+from transformers.pytorch_utils import ALL_LAYERNORM_LAYERS
+from transformers.utils import (TransformersKwargs, auto_docstring,
+                                can_return_tuple, logging)
+from transformers.utils.generic import OutputRecorder, check_model_inputs
+
+try:
+    from fla.layers.utils import get_unpad_data, index_first_axis, pad_input
+    from fla.modules import FusedRMSNormGated, ShortConvolution
+    from fla.ops.kda import chunk_kda, fused_recurrent_kda
+    from fla.ops.kda.gate import fused_kda_gate
+except ImportError:
+    raise ImportError("Plese run `pip install -U fla-core`")
+
+from .configuration_kimi import KimiLinearConfig
+
+assert version.parse(transformers.__version__) >= version.parse("4.56.0"), \
+    "Please upgrade transformers to >= 4.56.0"
+
+logger = logging.get_logger(__name__)
+
+
+class KimiDynamicCache:
+    """
+    Dynamic cache for Kimi model.
+    Inspired by Qwen3-Next
+    """
+    is_compileable = False
+
+    def __init__(self, config: KimiLinearConfig):
+        super().__init__()
+        self.config = config
+
+        if config.linear_attn_config is not None:
+            self.layer_types = []
+            for i in range(config.num_hidden_layers):
+                if config.is_kda_layer(i):
+                    self.layer_types.append("linear_attention")
+                else:
+                    self.layer_types.append("full_attention")
+        else:
+            self.layer_types = ["full_attention"] * config.num_hidden_layers
+
+        self.transformer_layers = [
+            i for i in range(config.num_hidden_layers) if self.layer_types[i] == "full_attention"
+        ]
+
+        linear_layers = [i for i in range(
+            config.num_hidden_layers) if self.layer_types[i] == "linear_attention"]
+        self.last_linear_layer = linear_layers[-1] if linear_layers else -1
+
+        self.conv_states = [None for _ in range(config.num_hidden_layers)]
+        self.recurrent_states = [None for _ in range(config.num_hidden_layers)]
+        self.key_cache = [None for _ in range(config.num_hidden_layers)]
+        self.value_cache = [None for _ in range(config.num_hidden_layers)]
+
+    def __len__(self):
+        return len(self.layer_types)
+
+    def update(
+        self,
+        key_states: torch.Tensor,
+        value_states: torch.Tensor,
+        layer_idx: int,
+        cache_kwargs: Optional[dict[str, Any]] = None,
+    ) -> tuple[torch.Tensor, torch.Tensor]:
+        if self.key_cache[layer_idx] is None:
+            self.key_cache[layer_idx] = key_states
+            self.value_cache[layer_idx] = value_states
+        else:
+            self.key_cache[layer_idx] = torch.cat(
+                [self.key_cache[layer_idx], key_states], dim=2)
+            self.value_cache[layer_idx] = torch.cat(
+                [self.value_cache[layer_idx], value_states], dim=2)
+
+        return self.key_cache[layer_idx], self.value_cache[layer_idx]
+
+    def reorder_cache(self, beam_idx: torch.LongTensor):
+        """Reorders the cache for beam search, given the selected beam indices."""
+        for layer_idx in range(len(self.key_cache)):
+            if self.key_cache[layer_idx] is not None:
+                device = self.key_cache[layer_idx].device
+                beam_idx = beam_idx.to(device)
+                self.key_cache[layer_idx] = self.key_cache[layer_idx].index_select(
+                    0, beam_idx)
+                self.value_cache[layer_idx] = self.value_cache[layer_idx].index_select(
+                    0, beam_idx)
+
+            if self.conv_states[layer_idx] is not None:
+                device = self.conv_states[layer_idx][0].device
+                beam_idx = beam_idx.to(device)
+                q_conv, k_conv, v_conv = self.conv_states[layer_idx]
+                self.conv_states[layer_idx] = (
+                    q_conv.index_select(0, beam_idx),
+                    k_conv.index_select(0, beam_idx),
+                    v_conv.index_select(0, beam_idx)
+                )
+                self.recurrent_states[layer_idx] = self.recurrent_states[layer_idx].index_select(
+                    0, beam_idx)
+
+    def get_seq_length(self, layer_idx: Optional[int] = 0) -> int:
+        """Returns the sequence length of the cached states. A layer index can be optionally passed."""
+        # take any layer that contains cache and not empty tensor
+        layer_idx = self.transformer_layers[0] if layer_idx not in self.transformer_layers else layer_idx
+        if len(self.key_cache) <= layer_idx or self.key_cache[layer_idx] is None:
+            return 0
+        return self.key_cache[layer_idx].shape[-2]
+
+    def get_mask_sizes(self, cache_position: torch.Tensor, layer_idx: int) -> tuple[int, int]:
+        """
+        Return a tuple (kv_length, kv_offset) corresponding to the length and offset that will be returned for
+        the given layer at `layer_idx`.
+        The masks are then prepared according to the given lengths (kv_length, kv_offset) and patterns for each layer.
+        """
+        kv_offset = 0
+        query_length = cache_position.shape[0]
+        past_seen_tokens = self.get_seq_length(layer_idx)
+        kv_length = query_length + past_seen_tokens
+        return kv_length, kv_offset
+
+    @property
+    def has_previous_state(self):
+        """We have a previous state if the last linear (conv) layer was already updated."""
+        if self.last_linear_layer == -1:
+            return False
+        return self.conv_states[self.last_linear_layer] is not None
+
+
+class KimiRMSNorm(nn.Module):
+    def __init__(self, hidden_size, eps=1e-6):
+        """
+        KimiRMSNorm is equivalent to T5LayerNorm
+        """
+        super().__init__()
+        self.weight = nn.Parameter(torch.ones(hidden_size))
+        self.variance_epsilon = eps
+
+    def forward(self, hidden_states):
+        input_dtype = hidden_states.dtype
+        hidden_states = hidden_states.to(torch.float32)
+        variance = hidden_states.pow(2).mean(-1, keepdim=True)
+        hidden_states = hidden_states * \
+            torch.rsqrt(variance + self.variance_epsilon)
+        return self.weight * hidden_states.to(input_dtype)
+
+
+ALL_LAYERNORM_LAYERS.append(KimiRMSNorm)
+
+
+class KimiBlockSparseMLP(nn.Module):
+    def __init__(self, config: KimiLinearConfig, hidden_size=None, intermediate_size=None):
+        super().__init__()
+        self.config = config
+        self.ffn_dim = config.intermediate_size if intermediate_size is None else intermediate_size
+        self.hidden_dim = config.hidden_size if hidden_size is None else hidden_size
+
+        self.w1 = nn.Linear(self.hidden_dim, self.ffn_dim, bias=False)   # gate
+        self.w2 = nn.Linear(self.ffn_dim, self.hidden_dim, bias=False)   # down
+        self.w3 = nn.Linear(self.hidden_dim, self.ffn_dim, bias=False)   # up
+
+        self.act_fn = ACT2FN[config.hidden_act]
+
+    def forward(self, hidden_states):
+        current_hidden_states = self.act_fn(
+            self.w1(hidden_states)) * self.w3(hidden_states)
+        current_hidden_states = self.w2(current_hidden_states)
+        return current_hidden_states
+
+
+class KimiMLP(nn.Module):
+    def __init__(self, config: KimiLinearConfig, hidden_size=None, intermediate_size=None):
+        super().__init__()
+        self.config = config
+        self.hidden_size = config.hidden_size if hidden_size is None else hidden_size
+        self.intermediate_size = config.intermediate_size if intermediate_size is None else intermediate_size
+        self.gate_proj = nn.Linear(
+            self.hidden_size, self.intermediate_size, bias=False)
+        self.up_proj = nn.Linear(
+            self.hidden_size, self.intermediate_size, bias=False)
+        self.down_proj = nn.Linear(
+            self.intermediate_size, self.hidden_size, bias=False)
+        self.act_fn = ACT2FN[config.hidden_act]
+
+    def forward(self, x):
+        down_proj = self.down_proj(self.act_fn(
+            self.gate_proj(x)) * self.up_proj(x))
+        return down_proj
+
+
+def repeat_kv(hidden_states: torch.Tensor, n_rep: int) -> torch.Tensor:
+    """
+    This is the equivalent of torch.repeat_interleave(x, dim=1, repeats=n_rep). The hidden states go from (batch,
+    num_key_value_heads, seqlen, head_dim) to (batch, num_attention_heads, seqlen, head_dim)
+    """
+    batch, num_key_value_heads, slen, head_dim = hidden_states.shape
+    if n_rep == 1:
+        return hidden_states
+    hidden_states = hidden_states[:, :, None, :, :].expand(
+        batch, num_key_value_heads, n_rep, slen, head_dim)
+    return hidden_states.reshape(batch, num_key_value_heads * n_rep, slen, head_dim)
+
+
+def eager_attention_forward(
+    module: nn.Module,
+    query: torch.Tensor,
+    key: torch.Tensor,
+    value: torch.Tensor,
+    attention_mask: Optional[torch.Tensor],
+    scaling: float,
+    dropout: float = 0.0,
+    **kwargs: Unpack[TransformersKwargs],
+):
+    key_states = repeat_kv(key, module.num_key_value_groups)
+    value_states = repeat_kv(value, module.num_key_value_groups)
+
+    attn_weights = torch.matmul(query, key_states.transpose(2, 3)) * scaling
+    if attention_mask is not None:
+        causal_mask = attention_mask[:, :, :, : key_states.shape[-2]]
+        attn_weights = attn_weights + causal_mask
+
+    attn_weights = nn.functional.softmax(
+        attn_weights, dim=-1, dtype=torch.float32).to(query.dtype)
+    attn_weights = nn.functional.dropout(
+        attn_weights, p=dropout, training=module.training)
+    attn_output = torch.matmul(attn_weights, value_states)
+    attn_output = attn_output.transpose(1, 2).contiguous()
+
+    return attn_output, attn_weights
+
+
+class KimiMLAAttention(nn.Module):
+    """
+    Multi-Latent Attention adapted from deepseek-v3
+    """
+
+    def __init__(self, config: KimiLinearConfig, layer_idx: int):
+        nn.Module.__init__(self)
+        self.config = config
+        self.layer_idx = layer_idx
+        self.hidden_size = config.hidden_size
+        self.num_heads = config.num_attention_heads
+        self.num_key_value_heads = config.num_key_value_heads
+        self.num_key_value_groups = self.num_heads // self.num_key_value_heads
+
+        self.rope_theta = config.rope_theta
+        self.attention_dropout = getattr(config, "attention_dropout", 0.0)
+
+        try:
+            self.q_lora_rank = config.q_lora_rank
+            self.qk_rope_head_dim = config.qk_rope_head_dim
+            self.kv_lora_rank = config.kv_lora_rank
+            self.v_head_dim = config.v_head_dim
+            self.qk_nope_head_dim = config.qk_nope_head_dim
+            self.q_head_dim = self.qk_nope_head_dim + self.qk_rope_head_dim
+            self.use_nope = config.mla_use_nope
+            self.scaling = self.q_head_dim ** (-0.5)
+        except Exception as e:
+            raise ValueError(
+                f"Kimi MLA config is not found or not properly formatted: {e}")
+
+        assert self.q_lora_rank is None
+        self.q_proj = nn.Linear(
+            self.hidden_size, self.num_heads * self.q_head_dim, bias=False,
+        )
+        self.kv_a_proj_with_mqa = nn.Linear(
+            self.hidden_size,
+            self.kv_lora_rank + self.qk_rope_head_dim,
+            bias=False,
+        )
+        self.kv_a_layernorm = KimiRMSNorm(self.kv_lora_rank)
+        self.kv_b_proj = nn.Linear(
+            self.kv_lora_rank,
+            self.num_heads
+            * (self.q_head_dim - self.qk_rope_head_dim + self.v_head_dim),
+            bias=False,
+        )
+        self.o_proj = nn.Linear(
+            self.num_heads * self.v_head_dim,
+            self.hidden_size,
+            bias=False,
+        )
+        self.is_causal = True
+        assert self.use_nope
+
+    def forward(
+        self,
+        hidden_states: torch.Tensor,
+        attention_mask: Optional[torch.Tensor] = None,
+        past_key_values: Optional[Cache] = None,
+        **kwargs,
+    ) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]:
+        batch_size, seq_length = hidden_states.shape[:-1]
+        query_shape = (batch_size, seq_length, -1, self.q_head_dim)
+        key_shape = (batch_size, seq_length, -1,
+                     self.qk_nope_head_dim + self.v_head_dim)
+
+        q_states = self.q_proj(hidden_states)
+        q_states = q_states.view(query_shape).transpose(1, 2)
+        q_pass, q_rot = torch.split(
+            q_states, [self.qk_nope_head_dim, self.qk_rope_head_dim], dim=-1)
+
+        compressed_kv = self.kv_a_proj_with_mqa(hidden_states)
+        k_pass, k_rot = torch.split(
+            compressed_kv, [self.kv_lora_rank, self.qk_rope_head_dim], dim=-1)
+
+        k_pass = self.kv_b_proj(self.kv_a_layernorm(
+            k_pass)).view(key_shape).transpose(1, 2)
+        k_pass, value_states = torch.split(
+            k_pass, [self.qk_nope_head_dim, self.v_head_dim], dim=-1)
+
+        k_rot = k_rot.view(batch_size, 1, seq_length, self.qk_rope_head_dim)
+        k_rot = k_rot.expand(*k_pass.shape[:-1], -1)
+
+        query_states = torch.cat((q_pass, q_rot), dim=-1)
+        key_states = torch.cat((k_pass, k_rot), dim=-1)
+
+        if past_key_values is not None:
+            key_states, value_states = past_key_values.update(
+                key_states, value_states, self.layer_idx)
+
+        if self.config._attn_implementation == "flash_attention_2" and self.q_head_dim != self.v_head_dim:
+            value_states = F.pad(
+                value_states, [0, self.q_head_dim - self.v_head_dim])
+
+        attention_interface: Callable = eager_attention_forward
+        if self.config._attn_implementation != "eager":
+            attention_interface = ALL_ATTENTION_FUNCTIONS[self.config._attn_implementation]
+
+        attn_output, _ = attention_interface(
+            self,
+            query_states,
+            key_states,
+            value_states,
+            attention_mask,
+            dropout=0.0 if not self.training else self.attention_dropout,
+            scaling=self.scaling,
+            **kwargs,
+        )
+
+        if self.config._attn_implementation == "flash_attention_2" and self.q_head_dim != self.v_head_dim:
+            attn_output = attn_output[:, :, :, : self.v_head_dim]
+
+        attn_output = attn_output.reshape(
+            batch_size, seq_length, -1).contiguous()
+        attn_output = self.o_proj(attn_output)
+        return attn_output
+
+
+class KimiDeltaAttention(nn.Module):
+    def __init__(self, config: KimiLinearConfig, layer_idx: int):
+        super().__init__()
+        self.config = config
+        self.mode = "chunk"
+
+        self.hidden_size = config.hidden_size
+        self.conv_size = config.linear_attn_config["short_conv_kernel_size"]
+        self.head_dim = config.linear_attn_config["head_dim"]
+        self.num_heads = config.linear_attn_config["num_heads"]
+        self.head_k_dim = self.head_dim
+        self.num_k_heads = self.num_heads
+
+        self.layer_idx = layer_idx
+
+        assert self.mode in [
+            'chunk', 'fused_recurrent'], f"Not suppoerted mode `{self.mode}`."
+
+        projection_k_size = self.head_k_dim * self.num_k_heads
+        projection_size = self.head_dim * self.num_heads
+
+        self.q_proj = nn.Linear(
+            self.hidden_size, projection_k_size, bias=False)
+        self.k_proj = nn.Linear(
+            self.hidden_size, projection_k_size, bias=False)
+        self.v_proj = nn.Linear(self.hidden_size, projection_size, bias=False)
+
+        self.q_conv1d = ShortConvolution(
+            hidden_size=projection_k_size,
+            kernel_size=self.conv_size,
+            activation='silu',
+        )
+        self.k_conv1d = ShortConvolution(
+            hidden_size=projection_k_size,
+            kernel_size=self.conv_size,
+            activation='silu'
+        )
+        self.v_conv1d = ShortConvolution(
+            hidden_size=projection_size,
+            kernel_size=self.conv_size,
+            activation='silu'
+        )
+
+        self.A_log = torch.nn.Parameter(torch.log(torch.empty(
+            self.num_heads, dtype=torch.float32).uniform_(1, 16)).view(1, 1, -1, 1))
+
+        self.f_a_proj = nn.Linear(self.hidden_size, self.head_dim, bias=False)
+        self.f_b_proj = nn.Linear(self.head_dim, projection_size, bias=False)
+
+        self.dt_bias = nn.Parameter(
+            torch.empty(projection_size, dtype=torch.float32))
+
+        self.b_proj = nn.Linear(self.hidden_size, self.num_heads, bias=False)
+
+        self.g_a_proj = nn.Linear(self.hidden_size, self.head_dim, bias=False)
+        self.g_b_proj = nn.Linear(self.head_dim, projection_size, bias=False)
+
+        self.o_norm = FusedRMSNormGated(
+            self.head_dim, eps=config.rms_norm_eps, activation='sigmoid')
+        self.o_proj = nn.Linear(projection_size, self.hidden_size, bias=False)
+
+    def forward(
+        self,
+        hidden_states: torch.Tensor,
+        attention_mask: Optional[torch.Tensor] = None,
+        cache_params: Optional[KimiDynamicCache] = None,
+        **kwargs: Unpack[dict]
+    ) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Cache]]:
+        if attention_mask is not None:
+            if attention_mask.dim() != 2:
+                attention_mask = kwargs.get("padding_mask", None)
+
+            if attention_mask is not None and attention_mask.dim() != 2:
+                raise ValueError(
+                    "attention_mask must be a 0-1 matrix of shape [batch_size, seq_len] "
+                    "(0 = padding). 3D masks are not supported here."
+                )
+        use_cache = cache_params is not None
+        batch_size, q_len, _ = hidden_states.shape
+        mode = 'fused_recurrent' if q_len <= 64 else self.mode
+        if self.training:
+            assert mode == 'chunk', "Only chunk mode is supported in training."
+
+        cu_seqlens = kwargs.get('cu_seqlens', None)
+        indices = None
+        if attention_mask is not None:
+            indices, cu_seqlens, _ = get_unpad_data(attention_mask[:, -q_len:])
+            hidden_states = index_first_axis(
+                rearrange(hidden_states, "b s ... -> (b s) ..."), indices).unsqueeze(0)
+
+        conv_state_q, conv_state_k, conv_state_v = None, None, None
+        recurrent_state = None
+        if cache_params is not None:
+            if cache_params.conv_states[self.layer_idx] is not None:
+                conv_state_q, conv_state_k, conv_state_v = cache_params.conv_states[
+                    self.layer_idx]
+            recurrent_state = cache_params.recurrent_states[self.layer_idx]
+        q, conv_state_q = self.q_conv1d(
+            x=self.q_proj(hidden_states),
+            cache=conv_state_q,
+            output_final_state=use_cache,
+            cu_seqlens=cu_seqlens
+        )
+        k, conv_state_k = self.k_conv1d(
+            x=self.k_proj(hidden_states),
+            cache=conv_state_k,
+            output_final_state=use_cache,
+            cu_seqlens=cu_seqlens
+        )
+        v, conv_state_v = self.v_conv1d(
+            x=self.v_proj(hidden_states),
+            cache=conv_state_v,
+            output_final_state=use_cache,
+            cu_seqlens=cu_seqlens
+        )
+        g = self.f_b_proj(self.f_a_proj(hidden_states))
+        g = fused_kda_gate(g, self.A_log, self.head_dim, g_bias=self.dt_bias)
+        beta = self.b_proj(hidden_states).float().sigmoid()
+
+        q, k = map(lambda x: rearrange(
+            x, '... (h d) -> ... h d', d=self.head_k_dim), (q, k))
+        v = rearrange(v, '... (h d) -> ... h d', d=self.head_dim)
+
+        if mode == 'chunk':
+            o, recurrent_state = chunk_kda(
+                q=q,
+                k=k,
+                v=v,
+                g=g,
+                beta=beta,
+                initial_state=recurrent_state,
+                output_final_state=True,
+                use_qk_l2norm_in_kernel=True,
+                cu_seqlens=cu_seqlens,
+            )
+        else:
+            o, recurrent_state = fused_recurrent_kda(
+                q=q,
+                k=k,
+                v=v,
+                g=g,
+                beta=beta,
+                initial_state=recurrent_state,
+                output_final_state=True,
+                use_qk_l2norm_in_kernel=True,
+                cu_seqlens=cu_seqlens,
+            )
+        if cache_params is not None:
+            cache_params.recurrent_states[self.layer_idx] = recurrent_state
+            cache_params.conv_states[self.layer_idx] = (
+                conv_state_q, conv_state_k, conv_state_v)
+
+        g = self.g_b_proj(self.g_a_proj(hidden_states))
+        g = rearrange(g, '... (h d) -> ... h d', d=self.head_dim)
+        o = self.o_norm(o, g)
+
+        o = rearrange(o, 'b t h d -> b t (h d)')
+        o = self.o_proj(o)
+        if attention_mask is not None:
+            o = pad_input(o.squeeze(0), indices, batch_size, q_len)
+
+        return o
+
+
+class KimiMoEGate(nn.Module):
+    """
+    MoEGate adapted from Deepseek-V3.
+    Parameter correspondences:
+        num_experts -> n_routed_experts
+        num_experts_per_token -> num_experts_per_tok
+        num_expert_group -> n_group
+        moe_router_activation_func -> scoring_func
+    """
+
+    def __init__(self, config: KimiLinearConfig):
+        super().__init__()
+        self.config = config
+        self.top_k = config.num_experts_per_token
+        self.num_experts = config.num_experts
+        self.routed_scaling_factor = config.routed_scaling_factor
+        self.moe_router_activation_func = config.moe_router_activation_func
+        self.num_expert_group = getattr(config, "num_expert_group", 1)
+        self.topk_group = getattr(config, "topk_group", 1)
+
+        # topk selection algorithm
+        self.moe_renormalize = config.moe_renormalize
+        self.gating_dim = config.hidden_size
+        self.weight = nn.Parameter(
+            torch.empty((self.num_experts, self.gating_dim))
+        )
+
+        self.e_score_correction_bias = nn.Parameter(
+            torch.empty((self.num_experts))
+        )
+        self.reset_parameters()
+
+    def reset_parameters(self) -> None:
+        import torch.nn.init as init
+
+        init.kaiming_uniform_(self.weight, a=math.sqrt(5))
+
+    def forward(self, hidden_states):
+        bsz, seq_len, h = hidden_states.shape
+        # compute gating score
+        hidden_states = hidden_states.view(-1, h)
+        logits = F.linear(
+            hidden_states.type(torch.float32), self.weight.type(
+                torch.float32), None
+        )
+        if self.moe_router_activation_func == "sigmoid":
+            scores = logits.sigmoid()
+        elif self.moe_router_activation_func == "softmax":
+            scores = logits.softmax(dim=1)
+        else:
+            raise NotImplementedError(
+                f"insupportable scoring function for MoE gating: {self.moe_router_activation_func}"
+            )
+
+        # select top-k experts
+        assert not self.training
+        scores_for_choice = scores.view(bsz * seq_len, -1)
+        scores_for_choice += self.e_score_correction_bias.unsqueeze(0)
+        group_scores = (
+            scores_for_choice.view(
+                bsz * seq_len, self.num_expert_group, -1).topk(2, dim=-1)[0].sum(dim=-1)
+        )  # [n, num_expert_group]
+        group_idx = torch.topk(
+            group_scores, k=self.topk_group, dim=-1, sorted=False
+        )[
+            1
+        ]  # [n, top_k_group]
+        group_mask = torch.zeros_like(group_scores)  # [n, num_expert_group]
+        group_mask.scatter_(1, group_idx, 1)  # [n, num_expert_group]
+        score_mask = (
+            group_mask.unsqueeze(-1)
+            .expand(
+                bsz * seq_len, self.num_expert_group, self.num_experts // self.num_expert_group
+            )
+            .reshape(bsz * seq_len, -1)
+        )  # [n, e]
+        tmp_scores = scores_for_choice.masked_fill(
+            ~score_mask.bool(), 0.0)  # [n, e]
+        _, topk_idx = torch.topk(
+            tmp_scores, k=self.top_k, dim=-1, sorted=False
+        )
+        topk_weight = scores.gather(1, topk_idx)
+
+        # norm gate to sum 1
+        if self.top_k > 1 and self.moe_renormalize:
+            denominator = topk_weight.sum(dim=-1, keepdim=True) + 1e-20
+            topk_weight = topk_weight / denominator
+        # must multiply the scaling factor
+        topk_weight = topk_weight * self.routed_scaling_factor
+
+        return topk_idx, topk_weight
+
+
+class KimiSparseMoeBlock(nn.Module):
+    """
+    Adapted from Deepseek-V3's MOE implementation
+    The namings are consistent with Kimi's version.
+    """
+
+    def __init__(self, config: KimiLinearConfig):
+        super().__init__()
+        self.config = config
+        self.hidden_dim = config.hidden_size
+        self.num_experts = config.num_experts
+        self.top_k = config.num_experts_per_token
+        self.moe_renormalize = config.moe_renormalize
+
+        self.ep_size = 1
+        self.experts_per_rank = config.num_experts
+        self.ep_rank = 0
+        self.experts = nn.ModuleList(
+            [
+                KimiBlockSparseMLP(
+                    config, intermediate_size=config.moe_intermediate_size
+                )
+                for _ in range(config.num_experts)
+            ]
+        )
+        self.gate = KimiMoEGate(config)
+        if config.num_shared_experts is not None:
+            intermediate_size = config.moe_intermediate_size * config.num_shared_experts
+            self.shared_experts = KimiMLP(
+                config=config, intermediate_size=intermediate_size
+            )
+
+    def forward(self, hidden_states):
+        identity = hidden_states
+        orig_shape = hidden_states.shape
+        topk_idx, topk_weight = self.gate(hidden_states)
+        hidden_states = hidden_states.view(-1, hidden_states.shape[-1])
+        flat_topk_idx = topk_idx.view(-1)
+        if not self.training:
+            y = self.moe_infer(hidden_states, topk_idx,
+                               topk_weight).view(*orig_shape)
+        else:
+            raise NotImplementedError(
+                "Training mode is not supported in KimiSparseMoeBlock")
+        if self.config.num_shared_experts is not None:
+            y = y + self.shared_experts(identity)
+        return y
+
+    @torch.no_grad()
+    def moe_infer(self, x, topk_ids, topk_weight):
+        cnts = topk_ids.new_zeros((topk_ids.shape[0], len(self.experts)))
+        cnts.scatter_(1, topk_ids, 1)
+        tokens_per_expert = cnts.sum(dim=0)
+        idxs = topk_ids.view(-1).argsort()
+        sorted_tokens = x[idxs // topk_ids.shape[1]]
+
+        tokens_per_expert = tokens_per_expert.cpu().numpy()
+
+        outputs = []
+        start_idx = 0
+        for i, num_tokens in enumerate(tokens_per_expert):
+            end_idx = start_idx + num_tokens
+            if num_tokens == 0:
+                continue
+            expert = self.experts[i + self.ep_rank * self.experts_per_rank]
+            tokens_for_this_expert = sorted_tokens[start_idx:end_idx]
+            expert_out = expert(tokens_for_this_expert)
+            outputs.append(expert_out)
+            start_idx = end_idx
+
+        outs = torch.cat(outputs, dim=0) if len(
+            outputs) else sorted_tokens.new_empty(0)
+
+        new_x = torch.empty_like(outs)
+        new_x[idxs] = outs
+        final_out = (
+            new_x.view(*topk_ids.shape, -1)
+            .type(topk_weight.dtype)
+            .mul_(topk_weight.unsqueeze(dim=-1))
+            .sum(dim=1)
+            .type(new_x.dtype)
+        )
+        return final_out
+
+
+class KimiDecoderLayer(nn.Module):
+    def __init__(self, config: KimiLinearConfig, layer_idx: int):
+        super().__init__()
+        self.hidden_size = config.hidden_size
+        self.config = config
+        if config.is_kda_layer(layer_idx):
+            self.is_linear_attn = True
+            self.self_attn = KimiDeltaAttention(
+                config=config, layer_idx=layer_idx)
+        elif config.is_mla:
+            self.is_linear_attn = False
+            self.self_attn = KimiMLAAttention(
+                config=config, layer_idx=layer_idx)
+        else:
+            raise NotImplementedError
+        if (
+            config.num_experts is not None
+            and layer_idx >= config.first_k_dense_replace
+            and layer_idx % getattr(config, "moe_layer_freq", 1) == 0
+        ):
+            self.block_sparse_moe = KimiSparseMoeBlock(config)
+        else:
+            self.mlp = KimiMLP(config)
+        self.input_layernorm = KimiRMSNorm(
+            config.hidden_size, eps=config.rms_norm_eps)
+        self.post_attention_layernorm = KimiRMSNorm(
+            config.hidden_size, eps=config.rms_norm_eps)
+
+    def forward(
+        self,
+        hidden_states: torch.Tensor,
+        attention_mask: Optional[torch.Tensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        past_key_values: Optional[Tuple[torch.Tensor]] = None,
+        output_attentions: Optional[bool] = False,
+        use_cache: Optional[bool] = False,
+        **kwargs: Unpack[FlashAttentionKwargs],
+    ) -> Tuple[torch.FloatTensor, Optional[Tuple[torch.FloatTensor, torch.FloatTensor]]]:
+        """
+        Args:
+            hidden_states (`torch.FloatTensor`): input to the layer of shape `(batch, seq_len, embed_dim)`
+            attention_mask (`torch.FloatTensor`, *optional*): attention mask of size
+                `(batch, 1, tgt_len, src_len)` where padding elements are indicated by very large negative values.
+            output_attentions (`bool`, *optional*):
+                Whether or not to return the attentions tensors of all attention layers. See `attentions` under
+                returned tensors for more detail.
+            use_cache (`bool`, *optional*):
+                If set to `True`, `past_key_values` key value states are returned and can be used to speed up decoding
+                (see `past_key_values`).
+            past_key_value (`Tuple(torch.FloatTensor)`, *optional*): cached past key and value projection states
+        """
+
+        residual = hidden_states
+
+        hidden_states = self.input_layernorm(hidden_states)
+
+        # Self Attention
+        if self.is_linear_attn is False:
+            hidden_states = self.self_attn(
+                hidden_states=hidden_states,
+                attention_mask=attention_mask,
+                position_ids=position_ids,
+                past_key_values=past_key_values,
+                output_attentions=output_attentions,
+                use_cache=use_cache,
+                **kwargs,
+            )
+        else:
+            hidden_states = self.self_attn(
+                hidden_states=hidden_states,
+                attention_mask=attention_mask,
+                cache_params=past_key_values,
+                output_attentions=output_attentions,
+                use_cache=use_cache,
+                **kwargs,
+            )
+        hidden_states = residual + hidden_states
+
+        # Fully Connected
+        residual = hidden_states
+        hidden_states = self.post_attention_layernorm(hidden_states)
+        if hasattr(self, "block_sparse_moe"):
+            hidden_states = self.block_sparse_moe(hidden_states)
+        else:
+            hidden_states = self.mlp(hidden_states)
+        hidden_states = residual + hidden_states
+
+        return hidden_states
+
+
+class KimiPreTrainedModel(PreTrainedModel):
+    config_class = KimiLinearConfig
+    base_model_prefix = "model"
+    supports_gradient_checkpointing = True
+    _no_split_modules = ["KimiDecoderLayer"]
+    _skip_keys_device_placement = "past_key_values"
+    _supports_flash_attn_2 = True
+    _can_record_outputs = {
+        "router_logits": OutputRecorder(KimiBlockSparseMLP, index=1),
+        "hidden_states": KimiDecoderLayer,
+        "attentions": KimiMLAAttention,
+    }
+    _is_stateful = True
+
+    def _init_weights(self, module):
+        std = self.config.initializer_range
+        if isinstance(module, nn.Linear):
+            module.weight.data.normal_(mean=0.0, std=std)
+            if module.bias is not None:
+                module.bias.data.zero_()
+        elif isinstance(module, nn.Embedding):
+            module.weight.data.normal_(mean=0.0, std=std)
+            if module.padding_idx is not None:
+                module.weight.data[module.padding_idx].zero_()
+
+
+class KimiLinearModel(KimiPreTrainedModel):
+    def __init__(self, config: KimiLinearConfig):
+        super().__init__(config)
+        self.padding_idx = config.pad_token_id
+        self.vocab_size = config.vocab_size
+
+        self.embed_tokens = nn.Embedding(
+            config.vocab_size, config.hidden_size, self.padding_idx)
+        self.layers = nn.ModuleList([KimiDecoderLayer(
+            config, layer_idx) for layer_idx in range(config.num_hidden_layers)])
+        self.norm = KimiRMSNorm(
+            config.hidden_size, eps=config.rms_norm_eps)
+
+        if getattr(config, "_attn_implementation", None) is not None:
+            if config._attn_implementation != "flash_attention_2":
+                logger.warning_once(
+                    f"Ignoring the provided attention implementation {config._attn_implementation}")
+                logger.warning_once("Using flash_attention_2 backend instead.")
+                config._attn_implementation = "flash_attention_2"
+        else:
+            config._attn_implementation = "flash_attention_2"
+
+        self._use_flash_attention_2 = config._attn_implementation == "flash_attention_2"
+        self.gradient_checkpointing = False
+        # Initialize weights and apply final processing
+        self.post_init()
+
+    def _update_linear_attn_mask(self, attention_mask, cache_position):
+        """
+        NOTE: Left-padding is used for linear attention mask.
+        No need for zeroing states when
+            1. Cached forward
+            2. Attending to all inputs
+        """
+        linear_attn_mask = attention_mask
+        if cache_position[0] > 0 or (attention_mask is not None and torch.all(attention_mask == 1)):
+            linear_attn_mask = None
+        return linear_attn_mask
+
+    @check_model_inputs
+    @auto_docstring
+    def forward(
+        self,
+        input_ids: torch.LongTensor = None,
+        attention_mask: Optional[torch.Tensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        past_key_values: Optional[Cache] = None,
+        inputs_embeds: Optional[torch.FloatTensor] = None,
+        cache_position: Optional[torch.LongTensor] = None,
+        use_cache: Optional[bool] = None,
+        **kwargs: Unpack[TransformersKwargs],
+    ) -> Union[Tuple, BaseModelOutputWithPast]:
+
+        use_cache = use_cache if use_cache is not None else self.config.use_cache
+
+        if (input_ids is None) and (inputs_embeds is None):
+            raise ValueError(
+                "You must specify exactly one of input_ids or inputs_embeds")
+
+        # Get inputs_embeds
+        if inputs_embeds is None:
+            inputs_embeds = self.embed_tokens(input_ids)
+
+        if use_cache and past_key_values is None:
+            past_key_values = KimiDynamicCache(config=self.config)
+
+        if cache_position is None:
+            past_seen_tokens = past_key_values.get_seq_length(
+            ) if past_key_values is not None else 0
+            cache_position: torch.Tensor = torch.arange(
+                past_seen_tokens, past_seen_tokens + inputs_embeds.shape[1], device=inputs_embeds.device
+            )
+
+        if position_ids is None:
+            position_ids = cache_position.unsqueeze(0)
+
+        causal_mask = create_causal_mask(
+            config=self.config,
+            input_embeds=inputs_embeds,
+            attention_mask=attention_mask,
+            cache_position=cache_position,
+            past_key_values=past_key_values,
+            position_ids=position_ids,
+        )
+        linear_attn_mask = self._update_linear_attn_mask(
+            attention_mask, cache_position)
+
+        hidden_states = inputs_embeds
+        if past_key_values is not None:
+            assert isinstance(past_key_values, KimiDynamicCache)
+
+        for decoder_layer in self.layers:
+            layer_mask = linear_attn_mask if decoder_layer.is_linear_attn else causal_mask
+
+            hidden_states = decoder_layer(
+                hidden_states,
+                attention_mask=layer_mask,
+                past_key_values=past_key_values,
+                cache_position=cache_position,
+                **kwargs,
+            )
+
+        hidden_states = self.norm(hidden_states)
+
+        return BaseModelOutputWithPast(
+            last_hidden_state=hidden_states,
+            past_key_values=past_key_values,
+        )
+
+
+class KimiLinearForCausalLM(KimiPreTrainedModel, GenerationMixin):
+    _tied_weights_keys = ["lm_head.weight"]
+
+    def __init__(self, config):
+        super().__init__(config)
+        self.model = KimiLinearModel(config)
+        self.vocab_size = config.vocab_size
+        self.lm_head = nn.Linear(
+            config.hidden_size, config.vocab_size, bias=False)
+
+        # Initialize weights and apply final processing
+        self.post_init()
+
+    @can_return_tuple
+    @auto_docstring
+    def forward(
+        self,
+        input_ids: torch.LongTensor = None,
+        attention_mask: Optional[torch.Tensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        past_key_values: Optional[List[torch.FloatTensor]] = None,
+        inputs_embeds: Optional[torch.FloatTensor] = None,
+        labels: Optional[torch.LongTensor] = None,
+        use_cache: Optional[bool] = None,
+        output_attentions: Optional[bool] = None,
+        output_hidden_states: Optional[bool] = None,
+        generation_mode: Optional[bool] = None,
+        return_dict: Optional[bool] = None,
+        cache_position: Optional[torch.LongTensor] = None,
+        **kwargs: Unpack[TransformersKwargs],
+    ) -> Union[Tuple, CausalLMOutputWithPast]:
+        r"""
+        Args:
+            labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
+                Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
+                config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored
+                (masked), the loss is only computed for the tokens with labels in `[0, ..., config.vocab_size]`.
+
+        Returns:
+
+        Example:
+
+        ```python
+        >>> from transformers import AutoTokenizer, KimiLinearForCausalLM
+
+        >>> model = KimiLinearForCausalLM.from_pretrained(PATH_TO_CONVERTED_WEIGHTS)
+        >>> tokenizer = AutoTokenizer.from_pretrained(PATH_TO_CONVERTED_TOKENIZER)
+
+        >>> prompt = "Hey, are you conscious? Can you talk to me?"
+        >>> inputs = tokenizer(prompt, return_tensors="pt")
+
+        >>> # Generate
+        >>> generate_ids = model.generate(inputs.input_ids, max_length=30)
+        >>> tokenizer.batch_decode(generate_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0]
+        "Hey, are you conscious? Can you talk to me?\nI'm not conscious, but I can talk to you."
+        ```"""
+
+        output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions
+        output_hidden_states = (
+            output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states
+        )
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+
+        outputs = self.model(
+            input_ids=input_ids,
+            attention_mask=attention_mask,
+            position_ids=position_ids,
+            past_key_values=past_key_values,
+            inputs_embeds=inputs_embeds,
+            use_cache=use_cache,
+            output_attentions=output_attentions,
+            output_hidden_states=output_hidden_states,
+            return_dict=return_dict,
+            cache_position=cache_position,
+        )
+
+        logits = outputs[0]
+        if generation_mode:
+            logits = logits[:, -1:]
+        logits = self.lm_head(logits)
+
+        loss = None
+        if labels is not None:
+            loss = self.loss_function(
+                logits, labels, self.vocab_size, **kwargs)
+
+        return CausalLMOutputWithPast(
+            loss=loss,
+            logits=logits,
+            past_key_values=outputs.past_key_values,
+            hidden_states=outputs.hidden_states,
+            attentions=outputs.attentions,
+        )