Upload folder using huggingface_hub

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:40d53fdb93a3abdf014f7683f4476d49b5527fc2d9af4fb0a76561ba22fa85a3
-size 5380457518
+oid sha256:daad2689d52002d444738ae649bfda64f23fc8f3857c39cf276ba80a6985d1fd
+size 5380457558

name_mapping.json

@@ -0,0 +1,228 @@
+{
+  "model.embed_tokens.weight": "embed.weight",
+  "model.hc_head.hc_base": "hc_head_base",
+  "model.hc_head.hc_fn": "hc_head_fn",
+  "model.hc_head.hc_scale": "hc_head_scale",
+  "model.layers.0.attn_hc.base": "layers.0.hc_attn_base",
+  "model.layers.0.attn_hc.fn": "layers.0.hc_attn_fn",
+  "model.layers.0.attn_hc.scale": "layers.0.hc_attn_scale",
+  "model.layers.0.ffn_hc.base": "layers.0.hc_ffn_base",
+  "model.layers.0.ffn_hc.fn": "layers.0.hc_ffn_fn",
+  "model.layers.0.ffn_hc.scale": "layers.0.hc_ffn_scale",
+  "model.layers.0.input_layernorm.weight": "layers.0.attn_norm.weight",
+  "model.layers.0.mlp.experts.0.w1.weight": "layers.0.ffn.experts.0.w1.weight",
+  "model.layers.0.mlp.experts.0.w2.weight": "layers.0.ffn.experts.0.w2.weight",
+  "model.layers.0.mlp.experts.0.w3.weight": "layers.0.ffn.experts.0.w3.weight",
+  "model.layers.0.mlp.experts.1.w1.weight": "layers.0.ffn.experts.1.w1.weight",
+  "model.layers.0.mlp.experts.1.w2.weight": "layers.0.ffn.experts.1.w2.weight",
+  "model.layers.0.mlp.experts.1.w3.weight": "layers.0.ffn.experts.1.w3.weight",
+  "model.layers.0.mlp.experts.10.w1.weight": "layers.0.ffn.experts.10.w1.weight",
+  "model.layers.0.mlp.experts.10.w2.weight": "layers.0.ffn.experts.10.w2.weight",
+  "model.layers.0.mlp.experts.10.w3.weight": "layers.0.ffn.experts.10.w3.weight",
+  "model.layers.0.mlp.experts.11.w1.weight": "layers.0.ffn.experts.11.w1.weight",
+  "model.layers.0.mlp.experts.11.w2.weight": "layers.0.ffn.experts.11.w2.weight",
+  "model.layers.0.mlp.experts.11.w3.weight": "layers.0.ffn.experts.11.w3.weight",
+  "model.layers.0.mlp.experts.12.w1.weight": "layers.0.ffn.experts.12.w1.weight",
+  "model.layers.0.mlp.experts.12.w2.weight": "layers.0.ffn.experts.12.w2.weight",
+  "model.layers.0.mlp.experts.12.w3.weight": "layers.0.ffn.experts.12.w3.weight",
+  "model.layers.0.mlp.experts.13.w1.weight": "layers.0.ffn.experts.13.w1.weight",
+  "model.layers.0.mlp.experts.13.w2.weight": "layers.0.ffn.experts.13.w2.weight",
+  "model.layers.0.mlp.experts.13.w3.weight": "layers.0.ffn.experts.13.w3.weight",
+  "model.layers.0.mlp.experts.14.w1.weight": "layers.0.ffn.experts.14.w1.weight",
+  "model.layers.0.mlp.experts.14.w2.weight": "layers.0.ffn.experts.14.w2.weight",
+  "model.layers.0.mlp.experts.14.w3.weight": "layers.0.ffn.experts.14.w3.weight",
+  "model.layers.0.mlp.experts.15.w1.weight": "layers.0.ffn.experts.15.w1.weight",
+  "model.layers.0.mlp.experts.15.w2.weight": "layers.0.ffn.experts.15.w2.weight",
+  "model.layers.0.mlp.experts.15.w3.weight": "layers.0.ffn.experts.15.w3.weight",
+  "model.layers.0.mlp.experts.2.w1.weight": "layers.0.ffn.experts.2.w1.weight",
+  "model.layers.0.mlp.experts.2.w2.weight": "layers.0.ffn.experts.2.w2.weight",
+  "model.layers.0.mlp.experts.2.w3.weight": "layers.0.ffn.experts.2.w3.weight",
+  "model.layers.0.mlp.experts.3.w1.weight": "layers.0.ffn.experts.3.w1.weight",
+  "model.layers.0.mlp.experts.3.w2.weight": "layers.0.ffn.experts.3.w2.weight",
+  "model.layers.0.mlp.experts.3.w3.weight": "layers.0.ffn.experts.3.w3.weight",
+  "model.layers.0.mlp.experts.4.w1.weight": "layers.0.ffn.experts.4.w1.weight",
+  "model.layers.0.mlp.experts.4.w2.weight": "layers.0.ffn.experts.4.w2.weight",
+  "model.layers.0.mlp.experts.4.w3.weight": "layers.0.ffn.experts.4.w3.weight",
+  "model.layers.0.mlp.experts.5.w1.weight": "layers.0.ffn.experts.5.w1.weight",
+  "model.layers.0.mlp.experts.5.w2.weight": "layers.0.ffn.experts.5.w2.weight",
+  "model.layers.0.mlp.experts.5.w3.weight": "layers.0.ffn.experts.5.w3.weight",
+  "model.layers.0.mlp.experts.6.w1.weight": "layers.0.ffn.experts.6.w1.weight",
+  "model.layers.0.mlp.experts.6.w2.weight": "layers.0.ffn.experts.6.w2.weight",
+  "model.layers.0.mlp.experts.6.w3.weight": "layers.0.ffn.experts.6.w3.weight",
+  "model.layers.0.mlp.experts.7.w1.weight": "layers.0.ffn.experts.7.w1.weight",
+  "model.layers.0.mlp.experts.7.w2.weight": "layers.0.ffn.experts.7.w2.weight",
+  "model.layers.0.mlp.experts.7.w3.weight": "layers.0.ffn.experts.7.w3.weight",
+  "model.layers.0.mlp.experts.8.w1.weight": "layers.0.ffn.experts.8.w1.weight",
+  "model.layers.0.mlp.experts.8.w2.weight": "layers.0.ffn.experts.8.w2.weight",
+  "model.layers.0.mlp.experts.8.w3.weight": "layers.0.ffn.experts.8.w3.weight",
+  "model.layers.0.mlp.experts.9.w1.weight": "layers.0.ffn.experts.9.w1.weight",
+  "model.layers.0.mlp.experts.9.w2.weight": "layers.0.ffn.experts.9.w2.weight",
+  "model.layers.0.mlp.experts.9.w3.weight": "layers.0.ffn.experts.9.w3.weight",
+  "model.layers.0.mlp.gate.tid2eid": "layers.0.ffn.gate.tid2eid",
+  "model.layers.0.mlp.gate.weight": "layers.0.ffn.gate.weight",
+  "model.layers.0.mlp.shared_experts.down_proj.weight": "layers.0.ffn.shared_experts.w2.weight",
+  "model.layers.0.mlp.shared_experts.gate_proj.weight": "layers.0.ffn.shared_experts.w1.weight",
+  "model.layers.0.mlp.shared_experts.up_proj.weight": "layers.0.ffn.shared_experts.w3.weight",
+  "model.layers.0.post_attention_layernorm.weight": "layers.0.ffn_norm.weight",
+  "model.layers.0.self_attn.compressor.gate_proj.weight": "layers.0.attn.compressor.gate_proj.weight",
+  "model.layers.0.self_attn.compressor.kv_norm.weight": "layers.0.attn.compressor.kv_norm.weight",
+  "model.layers.0.self_attn.compressor.kv_proj.weight": "layers.0.attn.compressor.kv_proj.weight",
+  "model.layers.0.self_attn.compressor.position_bias": "layers.0.attn.compressor.position_bias",
+  "model.layers.0.self_attn.kv_norm.weight": "layers.0.attn.kv_norm.weight",
+  "model.layers.0.self_attn.kv_proj.weight": "layers.0.attn.wkv.weight",
+  "model.layers.0.self_attn.o_a_proj.weight": "layers.0.attn.wo_a.weight",
+  "model.layers.0.self_attn.o_b_proj.weight": "layers.0.attn.wo_b.weight",
+  "model.layers.0.self_attn.q_a_norm.weight": "layers.0.attn.q_norm.weight",
+  "model.layers.0.self_attn.q_a_proj.weight": "layers.0.attn.wq_a.weight",
+  "model.layers.0.self_attn.q_b_proj.weight": "layers.0.attn.wq_b.weight",
+  "model.layers.0.self_attn.sinks": "layers.0.attn.attn_sink",
+  "model.layers.1.attn_hc.base": "layers.1.hc_attn_base",
+  "model.layers.1.attn_hc.fn": "layers.1.hc_attn_fn",
+  "model.layers.1.attn_hc.scale": "layers.1.hc_attn_scale",
+  "model.layers.1.ffn_hc.base": "layers.1.hc_ffn_base",
+  "model.layers.1.ffn_hc.fn": "layers.1.hc_ffn_fn",
+  "model.layers.1.ffn_hc.scale": "layers.1.hc_ffn_scale",
+  "model.layers.1.input_layernorm.weight": "layers.1.attn_norm.weight",
+  "model.layers.1.mlp.experts.0.w1.weight": "layers.1.ffn.experts.0.w1.weight",
+  "model.layers.1.mlp.experts.0.w2.weight": "layers.1.ffn.experts.0.w2.weight",
+  "model.layers.1.mlp.experts.0.w3.weight": "layers.1.ffn.experts.0.w3.weight",
+  "model.layers.1.mlp.experts.1.w1.weight": "layers.1.ffn.experts.1.w1.weight",
+  "model.layers.1.mlp.experts.1.w2.weight": "layers.1.ffn.experts.1.w2.weight",
+  "model.layers.1.mlp.experts.1.w3.weight": "layers.1.ffn.experts.1.w3.weight",
+  "model.layers.1.mlp.experts.10.w1.weight": "layers.1.ffn.experts.10.w1.weight",
+  "model.layers.1.mlp.experts.10.w2.weight": "layers.1.ffn.experts.10.w2.weight",
+  "model.layers.1.mlp.experts.10.w3.weight": "layers.1.ffn.experts.10.w3.weight",
+  "model.layers.1.mlp.experts.11.w1.weight": "layers.1.ffn.experts.11.w1.weight",
+  "model.layers.1.mlp.experts.11.w2.weight": "layers.1.ffn.experts.11.w2.weight",
+  "model.layers.1.mlp.experts.11.w3.weight": "layers.1.ffn.experts.11.w3.weight",
+  "model.layers.1.mlp.experts.12.w1.weight": "layers.1.ffn.experts.12.w1.weight",
+  "model.layers.1.mlp.experts.12.w2.weight": "layers.1.ffn.experts.12.w2.weight",
+  "model.layers.1.mlp.experts.12.w3.weight": "layers.1.ffn.experts.12.w3.weight",
+  "model.layers.1.mlp.experts.13.w1.weight": "layers.1.ffn.experts.13.w1.weight",
+  "model.layers.1.mlp.experts.13.w2.weight": "layers.1.ffn.experts.13.w2.weight",
+  "model.layers.1.mlp.experts.13.w3.weight": "layers.1.ffn.experts.13.w3.weight",
+  "model.layers.1.mlp.experts.14.w1.weight": "layers.1.ffn.experts.14.w1.weight",
+  "model.layers.1.mlp.experts.14.w2.weight": "layers.1.ffn.experts.14.w2.weight",
+  "model.layers.1.mlp.experts.14.w3.weight": "layers.1.ffn.experts.14.w3.weight",
+  "model.layers.1.mlp.experts.15.w1.weight": "layers.1.ffn.experts.15.w1.weight",
+  "model.layers.1.mlp.experts.15.w2.weight": "layers.1.ffn.experts.15.w2.weight",
+  "model.layers.1.mlp.experts.15.w3.weight": "layers.1.ffn.experts.15.w3.weight",
+  "model.layers.1.mlp.experts.2.w1.weight": "layers.1.ffn.experts.2.w1.weight",
+  "model.layers.1.mlp.experts.2.w2.weight": "layers.1.ffn.experts.2.w2.weight",
+  "model.layers.1.mlp.experts.2.w3.weight": "layers.1.ffn.experts.2.w3.weight",
+  "model.layers.1.mlp.experts.3.w1.weight": "layers.1.ffn.experts.3.w1.weight",
+  "model.layers.1.mlp.experts.3.w2.weight": "layers.1.ffn.experts.3.w2.weight",
+  "model.layers.1.mlp.experts.3.w3.weight": "layers.1.ffn.experts.3.w3.weight",
+  "model.layers.1.mlp.experts.4.w1.weight": "layers.1.ffn.experts.4.w1.weight",
+  "model.layers.1.mlp.experts.4.w2.weight": "layers.1.ffn.experts.4.w2.weight",
+  "model.layers.1.mlp.experts.4.w3.weight": "layers.1.ffn.experts.4.w3.weight",
+  "model.layers.1.mlp.experts.5.w1.weight": "layers.1.ffn.experts.5.w1.weight",
+  "model.layers.1.mlp.experts.5.w2.weight": "layers.1.ffn.experts.5.w2.weight",
+  "model.layers.1.mlp.experts.5.w3.weight": "layers.1.ffn.experts.5.w3.weight",
+  "model.layers.1.mlp.experts.6.w1.weight": "layers.1.ffn.experts.6.w1.weight",
+  "model.layers.1.mlp.experts.6.w2.weight": "layers.1.ffn.experts.6.w2.weight",
+  "model.layers.1.mlp.experts.6.w3.weight": "layers.1.ffn.experts.6.w3.weight",
+  "model.layers.1.mlp.experts.7.w1.weight": "layers.1.ffn.experts.7.w1.weight",
+  "model.layers.1.mlp.experts.7.w2.weight": "layers.1.ffn.experts.7.w2.weight",
+  "model.layers.1.mlp.experts.7.w3.weight": "layers.1.ffn.experts.7.w3.weight",
+  "model.layers.1.mlp.experts.8.w1.weight": "layers.1.ffn.experts.8.w1.weight",
+  "model.layers.1.mlp.experts.8.w2.weight": "layers.1.ffn.experts.8.w2.weight",
+  "model.layers.1.mlp.experts.8.w3.weight": "layers.1.ffn.experts.8.w3.weight",
+  "model.layers.1.mlp.experts.9.w1.weight": "layers.1.ffn.experts.9.w1.weight",
+  "model.layers.1.mlp.experts.9.w2.weight": "layers.1.ffn.experts.9.w2.weight",
+  "model.layers.1.mlp.experts.9.w3.weight": "layers.1.ffn.experts.9.w3.weight",
+  "model.layers.1.mlp.gate.e_score_correction_bias": "layers.1.ffn.gate.e_score_correction_bias",
+  "model.layers.1.mlp.gate.weight": "layers.1.ffn.gate.weight",
+  "model.layers.1.mlp.shared_experts.down_proj.weight": "layers.1.ffn.shared_experts.w2.weight",
+  "model.layers.1.mlp.shared_experts.gate_proj.weight": "layers.1.ffn.shared_experts.w1.weight",
+  "model.layers.1.mlp.shared_experts.up_proj.weight": "layers.1.ffn.shared_experts.w3.weight",
+  "model.layers.1.post_attention_layernorm.weight": "layers.1.ffn_norm.weight",
+  "model.layers.1.self_attn.compressor.gate_proj.weight": "layers.1.attn.compressor.gate_proj.weight",
+  "model.layers.1.self_attn.compressor.indexer.gate_proj.weight": "layers.1.attn.compressor.indexer.gate_proj.weight",
+  "model.layers.1.self_attn.compressor.indexer.kv_norm.weight": "layers.1.attn.compressor.indexer.kv_norm.weight",
+  "model.layers.1.self_attn.compressor.indexer.kv_proj.weight": "layers.1.attn.compressor.indexer.kv_proj.weight",
+  "model.layers.1.self_attn.compressor.indexer.position_bias": "layers.1.attn.compressor.indexer.position_bias",
+  "model.layers.1.self_attn.compressor.indexer.q_b_proj.weight": "layers.1.attn.compressor.indexer.q_b_proj.weight",
+  "model.layers.1.self_attn.compressor.indexer.weights_proj.weight": "layers.1.attn.compressor.indexer.weights_proj.weight",
+  "model.layers.1.self_attn.compressor.kv_norm.weight": "layers.1.attn.compressor.kv_norm.weight",
+  "model.layers.1.self_attn.compressor.kv_proj.weight": "layers.1.attn.compressor.kv_proj.weight",
+  "model.layers.1.self_attn.compressor.position_bias": "layers.1.attn.compressor.position_bias",
+  "model.layers.1.self_attn.kv_norm.weight": "layers.1.attn.kv_norm.weight",
+  "model.layers.1.self_attn.kv_proj.weight": "layers.1.attn.wkv.weight",
+  "model.layers.1.self_attn.o_a_proj.weight": "layers.1.attn.wo_a.weight",
+  "model.layers.1.self_attn.o_b_proj.weight": "layers.1.attn.wo_b.weight",
+  "model.layers.1.self_attn.q_a_norm.weight": "layers.1.attn.q_norm.weight",
+  "model.layers.1.self_attn.q_a_proj.weight": "layers.1.attn.wq_a.weight",
+  "model.layers.1.self_attn.q_b_proj.weight": "layers.1.attn.wq_b.weight",
+  "model.layers.1.self_attn.sinks": "layers.1.attn.attn_sink",
+  "model.layers.2.attn_hc.base": "layers.2.hc_attn_base",
+  "model.layers.2.attn_hc.fn": "layers.2.hc_attn_fn",
+  "model.layers.2.attn_hc.scale": "layers.2.hc_attn_scale",
+  "model.layers.2.ffn_hc.base": "layers.2.hc_ffn_base",
+  "model.layers.2.ffn_hc.fn": "layers.2.hc_ffn_fn",
+  "model.layers.2.ffn_hc.scale": "layers.2.hc_ffn_scale",
+  "model.layers.2.input_layernorm.weight": "layers.2.attn_norm.weight",
+  "model.layers.2.mlp.experts.0.w1.weight": "layers.2.ffn.experts.0.w1.weight",
+  "model.layers.2.mlp.experts.0.w2.weight": "layers.2.ffn.experts.0.w2.weight",
+  "model.layers.2.mlp.experts.0.w3.weight": "layers.2.ffn.experts.0.w3.weight",
+  "model.layers.2.mlp.experts.1.w1.weight": "layers.2.ffn.experts.1.w1.weight",
+  "model.layers.2.mlp.experts.1.w2.weight": "layers.2.ffn.experts.1.w2.weight",
+  "model.layers.2.mlp.experts.1.w3.weight": "layers.2.ffn.experts.1.w3.weight",
+  "model.layers.2.mlp.experts.10.w1.weight": "layers.2.ffn.experts.10.w1.weight",
+  "model.layers.2.mlp.experts.10.w2.weight": "layers.2.ffn.experts.10.w2.weight",
+  "model.layers.2.mlp.experts.10.w3.weight": "layers.2.ffn.experts.10.w3.weight",
+  "model.layers.2.mlp.experts.11.w1.weight": "layers.2.ffn.experts.11.w1.weight",
+  "model.layers.2.mlp.experts.11.w2.weight": "layers.2.ffn.experts.11.w2.weight",
+  "model.layers.2.mlp.experts.11.w3.weight": "layers.2.ffn.experts.11.w3.weight",
+  "model.layers.2.mlp.experts.12.w1.weight": "layers.2.ffn.experts.12.w1.weight",
+  "model.layers.2.mlp.experts.12.w2.weight": "layers.2.ffn.experts.12.w2.weight",
+  "model.layers.2.mlp.experts.12.w3.weight": "layers.2.ffn.experts.12.w3.weight",
+  "model.layers.2.mlp.experts.13.w1.weight": "layers.2.ffn.experts.13.w1.weight",
+  "model.layers.2.mlp.experts.13.w2.weight": "layers.2.ffn.experts.13.w2.weight",
+  "model.layers.2.mlp.experts.13.w3.weight": "layers.2.ffn.experts.13.w3.weight",
+  "model.layers.2.mlp.experts.14.w1.weight": "layers.2.ffn.experts.14.w1.weight",
+  "model.layers.2.mlp.experts.14.w2.weight": "layers.2.ffn.experts.14.w2.weight",
+  "model.layers.2.mlp.experts.14.w3.weight": "layers.2.ffn.experts.14.w3.weight",
+  "model.layers.2.mlp.experts.15.w1.weight": "layers.2.ffn.experts.15.w1.weight",
+  "model.layers.2.mlp.experts.15.w2.weight": "layers.2.ffn.experts.15.w2.weight",
+  "model.layers.2.mlp.experts.15.w3.weight": "layers.2.ffn.experts.15.w3.weight",
+  "model.layers.2.mlp.experts.2.w1.weight": "layers.2.ffn.experts.2.w1.weight",
+  "model.layers.2.mlp.experts.2.w2.weight": "layers.2.ffn.experts.2.w2.weight",
+  "model.layers.2.mlp.experts.2.w3.weight": "layers.2.ffn.experts.2.w3.weight",
+  "model.layers.2.mlp.experts.3.w1.weight": "layers.2.ffn.experts.3.w1.weight",
+  "model.layers.2.mlp.experts.3.w2.weight": "layers.2.ffn.experts.3.w2.weight",
+  "model.layers.2.mlp.experts.3.w3.weight": "layers.2.ffn.experts.3.w3.weight",
+  "model.layers.2.mlp.experts.4.w1.weight": "layers.2.ffn.experts.4.w1.weight",
+  "model.layers.2.mlp.experts.4.w2.weight": "layers.2.ffn.experts.4.w2.weight",
+  "model.layers.2.mlp.experts.4.w3.weight": "layers.2.ffn.experts.4.w3.weight",
+  "model.layers.2.mlp.experts.5.w1.weight": "layers.2.ffn.experts.5.w1.weight",
+  "model.layers.2.mlp.experts.5.w2.weight": "layers.2.ffn.experts.5.w2.weight",
+  "model.layers.2.mlp.experts.5.w3.weight": "layers.2.ffn.experts.5.w3.weight",
+  "model.layers.2.mlp.experts.6.w1.weight": "layers.2.ffn.experts.6.w1.weight",
+  "model.layers.2.mlp.experts.6.w2.weight": "layers.2.ffn.experts.6.w2.weight",
+  "model.layers.2.mlp.experts.6.w3.weight": "layers.2.ffn.experts.6.w3.weight",
+  "model.layers.2.mlp.experts.7.w1.weight": "layers.2.ffn.experts.7.w1.weight",
+  "model.layers.2.mlp.experts.7.w2.weight": "layers.2.ffn.experts.7.w2.weight",
+  "model.layers.2.mlp.experts.7.w3.weight": "layers.2.ffn.experts.7.w3.weight",
+  "model.layers.2.mlp.experts.8.w1.weight": "layers.2.ffn.experts.8.w1.weight",
+  "model.layers.2.mlp.experts.8.w2.weight": "layers.2.ffn.experts.8.w2.weight",
+  "model.layers.2.mlp.experts.8.w3.weight": "layers.2.ffn.experts.8.w3.weight",
+  "model.layers.2.mlp.experts.9.w1.weight": "layers.2.ffn.experts.9.w1.weight",
+  "model.layers.2.mlp.experts.9.w2.weight": "layers.2.ffn.experts.9.w2.weight",
+  "model.layers.2.mlp.experts.9.w3.weight": "layers.2.ffn.experts.9.w3.weight",
+  "model.layers.2.mlp.gate.e_score_correction_bias": "layers.2.ffn.gate.e_score_correction_bias",
+  "model.layers.2.mlp.gate.weight": "layers.2.ffn.gate.weight",
+  "model.layers.2.mlp.shared_experts.down_proj.weight": "layers.2.ffn.shared_experts.w2.weight",
+  "model.layers.2.mlp.shared_experts.gate_proj.weight": "layers.2.ffn.shared_experts.w1.weight",
+  "model.layers.2.mlp.shared_experts.up_proj.weight": "layers.2.ffn.shared_experts.w3.weight",
+  "model.layers.2.post_attention_layernorm.weight": "layers.2.ffn_norm.weight",
+  "model.layers.2.self_attn.kv_norm.weight": "layers.2.attn.kv_norm.weight",
+  "model.layers.2.self_attn.kv_proj.weight": "layers.2.attn.wkv.weight",
+  "model.layers.2.self_attn.o_a_proj.weight": "layers.2.attn.wo_a.weight",
+  "model.layers.2.self_attn.o_b_proj.weight": "layers.2.attn.wo_b.weight",
+  "model.layers.2.self_attn.q_a_norm.weight": "layers.2.attn.q_norm.weight",
+  "model.layers.2.self_attn.q_a_proj.weight": "layers.2.attn.wq_a.weight",
+  "model.layers.2.self_attn.q_b_proj.weight": "layers.2.attn.wq_b.weight",
+  "model.layers.2.self_attn.sinks": "layers.2.attn.attn_sink",
+  "model.norm.weight": "norm.weight"
+}