qwen3_5_4b_base / model.weights.json
prasadsachin's picture
Upload folder using huggingface_hub
2b5f3ff verified
Raw
History Blame Contribute Delete
60.4 kB
{
"metadata": {
"total_size": 18157062144.0
},
"weight_map": {
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block/attn/proj/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block/attn/qkv/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block/mlp/fc1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block/mlp/fc2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block/norm1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block/norm2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_1/attn/proj/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_1/attn/qkv/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_1/mlp/fc1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_1/mlp/fc2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_1/norm1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_1/norm2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_2/attn/proj/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_2/attn/qkv/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_2/mlp/fc1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_2/mlp/fc2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_2/norm1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_2/norm2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_3/attn/proj/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_3/attn/qkv/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_3/mlp/fc1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_3/mlp/fc2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_3/norm1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_3/norm2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_4/attn/proj/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_4/attn/qkv/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_4/mlp/fc1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_4/mlp/fc2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_4/norm1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_4/norm2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_5/attn/proj/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_5/attn/qkv/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_5/mlp/fc1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_5/mlp/fc2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_5/norm1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_5/norm2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_6/attn/proj/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_6/attn/qkv/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_6/mlp/fc1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_6/mlp/fc2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_6/norm1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_6/norm2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_7/attn/proj/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_7/attn/qkv/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_7/mlp/fc1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_7/mlp/fc2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_7/norm1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_7/norm2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_8/attn/proj/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_8/attn/qkv/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_8/mlp/fc1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_8/mlp/fc2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_8/norm1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_8/norm2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_9/attn/proj/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_9/attn/qkv/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_9/mlp/fc1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_9/mlp/fc2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_9/norm1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_9/norm2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_10/attn/proj/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_10/attn/qkv/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_10/mlp/fc1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_10/mlp/fc2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_10/norm1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_10/norm2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_11/attn/proj/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_11/attn/qkv/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_11/mlp/fc1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_11/mlp/fc2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_11/norm1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_11/norm2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_12/attn/proj/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_12/attn/qkv/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_12/mlp/fc1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_12/mlp/fc2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_12/norm1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_12/norm2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_13/attn/proj/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_13/attn/qkv/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_13/mlp/fc1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_13/mlp/fc2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_13/norm1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_13/norm2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_14/attn/proj/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_14/attn/qkv/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_14/mlp/fc1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_14/mlp/fc2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_14/norm1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_14/norm2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_15/attn/proj/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_15/attn/qkv/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_15/mlp/fc1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_15/mlp/fc2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_15/norm1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_15/norm2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_16/attn/proj/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_16/attn/qkv/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_16/mlp/fc1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_16/mlp/fc2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_16/norm1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_16/norm2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_17/attn/proj/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_17/attn/qkv/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_17/mlp/fc1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_17/mlp/fc2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_17/norm1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_17/norm2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_18/attn/proj/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_18/attn/qkv/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_18/mlp/fc1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_18/mlp/fc2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_18/norm1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_18/norm2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_19/attn/proj/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_19/attn/qkv/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_19/mlp/fc1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_19/mlp/fc2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_19/norm1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_19/norm2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_20/attn/proj/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_20/attn/qkv/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_20/mlp/fc1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_20/mlp/fc2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_20/norm1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_20/norm2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_21/attn/proj/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_21/attn/qkv/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_21/mlp/fc1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_21/mlp/fc2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_21/norm1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_21/norm2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_22/attn/proj/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_22/attn/qkv/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_22/mlp/fc1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_22/mlp/fc2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_22/norm1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_22/norm2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_23/attn/proj/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_23/attn/qkv/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_23/mlp/fc1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_23/mlp/fc2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_23/norm1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/blocks/qwen3_5_vision_block_23/norm2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/layers/qwen3_5_vision_patch_embed/proj/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/layers/embedding/vars": "model_00000.weights.h5",
"/layers/qwen3_5_vision_encoder/layers/qwen3_5_vision_patch_merger/fc1/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/layers/qwen3_5_vision_patch_merger/fc2/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_vision_encoder/layers/qwen3_5_vision_patch_merger/norm/vars": [
"model_00000.weights.h5"
],
"/layers/reversible_embedding/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder/_feedforward_gate_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder/_feedforward_intermediate_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder/_feedforward_output_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder/_input_layernorm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder/_linear_attn/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_transformer_decoder/_linear_attn/in_proj_a/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder/_linear_attn/in_proj_b/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder/_linear_attn/in_proj_qkv/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder/_linear_attn/in_proj_z/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder/_linear_attn/norm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder/_linear_attn/out_proj/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder/_post_attention_layernorm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_1/_feedforward_gate_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_1/_feedforward_intermediate_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_1/_feedforward_output_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_1/_input_layernorm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_1/_linear_attn/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_transformer_decoder_1/_linear_attn/in_proj_a/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_1/_linear_attn/in_proj_b/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_1/_linear_attn/in_proj_qkv/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_1/_linear_attn/in_proj_z/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_1/_linear_attn/norm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_1/_linear_attn/out_proj/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_1/_post_attention_layernorm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_2/_feedforward_gate_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_2/_feedforward_intermediate_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_2/_feedforward_output_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_2/_input_layernorm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_2/_linear_attn/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_transformer_decoder_2/_linear_attn/in_proj_a/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_2/_linear_attn/in_proj_b/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_2/_linear_attn/in_proj_qkv/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_2/_linear_attn/in_proj_z/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_2/_linear_attn/norm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_2/_linear_attn/out_proj/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_2/_post_attention_layernorm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_3/_feedforward_gate_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_3/_feedforward_intermediate_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_3/_feedforward_output_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_3/_input_layernorm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_3/_post_attention_layernorm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_3/_self_attention_layer/_key_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_3/_self_attention_layer/_key_norm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_3/_self_attention_layer/_output_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_3/_self_attention_layer/_query_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_3/_self_attention_layer/_query_norm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_3/_self_attention_layer/_value_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_4/_feedforward_gate_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_4/_feedforward_intermediate_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_4/_feedforward_output_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_4/_input_layernorm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_4/_linear_attn/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_transformer_decoder_4/_linear_attn/in_proj_a/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_4/_linear_attn/in_proj_b/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_4/_linear_attn/in_proj_qkv/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_4/_linear_attn/in_proj_z/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_4/_linear_attn/norm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_4/_linear_attn/out_proj/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_4/_post_attention_layernorm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_5/_feedforward_gate_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_5/_feedforward_intermediate_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_5/_feedforward_output_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_5/_input_layernorm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_5/_linear_attn/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_transformer_decoder_5/_linear_attn/in_proj_a/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_5/_linear_attn/in_proj_b/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_5/_linear_attn/in_proj_qkv/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_5/_linear_attn/in_proj_z/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_5/_linear_attn/norm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_5/_linear_attn/out_proj/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_5/_post_attention_layernorm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_6/_feedforward_gate_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_6/_feedforward_intermediate_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_6/_feedforward_output_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_6/_input_layernorm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_6/_linear_attn/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_transformer_decoder_6/_linear_attn/in_proj_a/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_6/_linear_attn/in_proj_b/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_6/_linear_attn/in_proj_qkv/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_6/_linear_attn/in_proj_z/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_6/_linear_attn/norm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_6/_linear_attn/out_proj/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_6/_post_attention_layernorm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_7/_feedforward_gate_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_7/_feedforward_intermediate_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_7/_feedforward_output_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_7/_input_layernorm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_7/_post_attention_layernorm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_7/_self_attention_layer/_key_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_7/_self_attention_layer/_key_norm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_7/_self_attention_layer/_output_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_7/_self_attention_layer/_query_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_7/_self_attention_layer/_query_norm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_7/_self_attention_layer/_value_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_8/_feedforward_gate_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_8/_feedforward_intermediate_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_8/_feedforward_output_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_8/_input_layernorm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_8/_linear_attn/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_transformer_decoder_8/_linear_attn/in_proj_a/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_8/_linear_attn/in_proj_b/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_8/_linear_attn/in_proj_qkv/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_8/_linear_attn/in_proj_z/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_8/_linear_attn/norm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_8/_linear_attn/out_proj/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_8/_post_attention_layernorm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_9/_feedforward_gate_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_9/_feedforward_intermediate_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_9/_feedforward_output_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_9/_input_layernorm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_9/_linear_attn/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_transformer_decoder_9/_linear_attn/in_proj_a/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_9/_linear_attn/in_proj_b/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_9/_linear_attn/in_proj_qkv/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_9/_linear_attn/in_proj_z/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_9/_linear_attn/norm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_9/_linear_attn/out_proj/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_9/_post_attention_layernorm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_10/_feedforward_gate_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_10/_feedforward_intermediate_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_10/_feedforward_output_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_10/_input_layernorm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_10/_linear_attn/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_transformer_decoder_10/_linear_attn/in_proj_a/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_10/_linear_attn/in_proj_b/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_10/_linear_attn/in_proj_qkv/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_10/_linear_attn/in_proj_z/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_10/_linear_attn/norm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_10/_linear_attn/out_proj/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_10/_post_attention_layernorm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_11/_feedforward_gate_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_11/_feedforward_intermediate_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_11/_feedforward_output_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_11/_input_layernorm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_11/_post_attention_layernorm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_11/_self_attention_layer/_key_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_11/_self_attention_layer/_key_norm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_11/_self_attention_layer/_output_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_11/_self_attention_layer/_query_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_11/_self_attention_layer/_query_norm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_11/_self_attention_layer/_value_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_12/_feedforward_gate_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_12/_feedforward_intermediate_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_12/_feedforward_output_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_12/_input_layernorm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_12/_linear_attn/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_transformer_decoder_12/_linear_attn/in_proj_a/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_12/_linear_attn/in_proj_b/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_12/_linear_attn/in_proj_qkv/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_12/_linear_attn/in_proj_z/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_12/_linear_attn/norm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_12/_linear_attn/out_proj/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_12/_post_attention_layernorm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_13/_feedforward_gate_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_13/_feedforward_intermediate_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_13/_feedforward_output_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_13/_input_layernorm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_13/_linear_attn/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_transformer_decoder_13/_linear_attn/in_proj_a/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_13/_linear_attn/in_proj_b/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_13/_linear_attn/in_proj_qkv/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_13/_linear_attn/in_proj_z/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_13/_linear_attn/norm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_13/_linear_attn/out_proj/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_13/_post_attention_layernorm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_14/_feedforward_gate_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_14/_feedforward_intermediate_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_14/_feedforward_output_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_14/_input_layernorm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_14/_linear_attn/vars": [
"model_00000.weights.h5"
],
"/layers/qwen3_5_transformer_decoder_14/_linear_attn/in_proj_a/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_14/_linear_attn/in_proj_b/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_14/_linear_attn/in_proj_qkv/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_14/_linear_attn/in_proj_z/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_14/_linear_attn/norm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_14/_linear_attn/out_proj/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_14/_post_attention_layernorm/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_15/_feedforward_gate_dense/vars": "model_00000.weights.h5",
"/layers/qwen3_5_transformer_decoder_15/_feedforward_intermediate_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_15/_feedforward_output_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_15/_input_layernorm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_15/_post_attention_layernorm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_15/_self_attention_layer/_key_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_15/_self_attention_layer/_key_norm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_15/_self_attention_layer/_output_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_15/_self_attention_layer/_query_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_15/_self_attention_layer/_query_norm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_15/_self_attention_layer/_value_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_16/_feedforward_gate_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_16/_feedforward_intermediate_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_16/_feedforward_output_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_16/_input_layernorm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_16/_linear_attn/vars": [
"model_00001.weights.h5"
],
"/layers/qwen3_5_transformer_decoder_16/_linear_attn/in_proj_a/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_16/_linear_attn/in_proj_b/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_16/_linear_attn/in_proj_qkv/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_16/_linear_attn/in_proj_z/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_16/_linear_attn/norm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_16/_linear_attn/out_proj/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_16/_post_attention_layernorm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_17/_feedforward_gate_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_17/_feedforward_intermediate_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_17/_feedforward_output_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_17/_input_layernorm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_17/_linear_attn/vars": [
"model_00001.weights.h5"
],
"/layers/qwen3_5_transformer_decoder_17/_linear_attn/in_proj_a/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_17/_linear_attn/in_proj_b/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_17/_linear_attn/in_proj_qkv/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_17/_linear_attn/in_proj_z/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_17/_linear_attn/norm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_17/_linear_attn/out_proj/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_17/_post_attention_layernorm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_18/_feedforward_gate_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_18/_feedforward_intermediate_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_18/_feedforward_output_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_18/_input_layernorm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_18/_linear_attn/vars": [
"model_00001.weights.h5"
],
"/layers/qwen3_5_transformer_decoder_18/_linear_attn/in_proj_a/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_18/_linear_attn/in_proj_b/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_18/_linear_attn/in_proj_qkv/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_18/_linear_attn/in_proj_z/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_18/_linear_attn/norm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_18/_linear_attn/out_proj/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_18/_post_attention_layernorm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_19/_feedforward_gate_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_19/_feedforward_intermediate_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_19/_feedforward_output_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_19/_input_layernorm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_19/_post_attention_layernorm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_19/_self_attention_layer/_key_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_19/_self_attention_layer/_key_norm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_19/_self_attention_layer/_output_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_19/_self_attention_layer/_query_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_19/_self_attention_layer/_query_norm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_19/_self_attention_layer/_value_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_20/_feedforward_gate_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_20/_feedforward_intermediate_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_20/_feedforward_output_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_20/_input_layernorm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_20/_linear_attn/vars": [
"model_00001.weights.h5"
],
"/layers/qwen3_5_transformer_decoder_20/_linear_attn/in_proj_a/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_20/_linear_attn/in_proj_b/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_20/_linear_attn/in_proj_qkv/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_20/_linear_attn/in_proj_z/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_20/_linear_attn/norm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_20/_linear_attn/out_proj/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_20/_post_attention_layernorm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_21/_feedforward_gate_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_21/_feedforward_intermediate_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_21/_feedforward_output_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_21/_input_layernorm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_21/_linear_attn/vars": [
"model_00001.weights.h5"
],
"/layers/qwen3_5_transformer_decoder_21/_linear_attn/in_proj_a/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_21/_linear_attn/in_proj_b/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_21/_linear_attn/in_proj_qkv/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_21/_linear_attn/in_proj_z/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_21/_linear_attn/norm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_21/_linear_attn/out_proj/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_21/_post_attention_layernorm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_22/_feedforward_gate_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_22/_feedforward_intermediate_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_22/_feedforward_output_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_22/_input_layernorm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_22/_linear_attn/vars": [
"model_00001.weights.h5"
],
"/layers/qwen3_5_transformer_decoder_22/_linear_attn/in_proj_a/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_22/_linear_attn/in_proj_b/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_22/_linear_attn/in_proj_qkv/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_22/_linear_attn/in_proj_z/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_22/_linear_attn/norm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_22/_linear_attn/out_proj/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_22/_post_attention_layernorm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_23/_feedforward_gate_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_23/_feedforward_intermediate_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_23/_feedforward_output_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_23/_input_layernorm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_23/_post_attention_layernorm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_23/_self_attention_layer/_key_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_23/_self_attention_layer/_key_norm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_23/_self_attention_layer/_output_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_23/_self_attention_layer/_query_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_23/_self_attention_layer/_query_norm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_23/_self_attention_layer/_value_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_24/_feedforward_gate_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_24/_feedforward_intermediate_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_24/_feedforward_output_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_24/_input_layernorm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_24/_linear_attn/vars": [
"model_00001.weights.h5"
],
"/layers/qwen3_5_transformer_decoder_24/_linear_attn/in_proj_a/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_24/_linear_attn/in_proj_b/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_24/_linear_attn/in_proj_qkv/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_24/_linear_attn/in_proj_z/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_24/_linear_attn/norm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_24/_linear_attn/out_proj/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_24/_post_attention_layernorm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_25/_feedforward_gate_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_25/_feedforward_intermediate_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_25/_feedforward_output_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_25/_input_layernorm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_25/_linear_attn/vars": [
"model_00001.weights.h5"
],
"/layers/qwen3_5_transformer_decoder_25/_linear_attn/in_proj_a/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_25/_linear_attn/in_proj_b/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_25/_linear_attn/in_proj_qkv/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_25/_linear_attn/in_proj_z/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_25/_linear_attn/norm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_25/_linear_attn/out_proj/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_25/_post_attention_layernorm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_26/_feedforward_gate_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_26/_feedforward_intermediate_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_26/_feedforward_output_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_26/_input_layernorm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_26/_linear_attn/vars": [
"model_00001.weights.h5"
],
"/layers/qwen3_5_transformer_decoder_26/_linear_attn/in_proj_a/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_26/_linear_attn/in_proj_b/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_26/_linear_attn/in_proj_qkv/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_26/_linear_attn/in_proj_z/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_26/_linear_attn/norm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_26/_linear_attn/out_proj/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_26/_post_attention_layernorm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_27/_feedforward_gate_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_27/_feedforward_intermediate_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_27/_feedforward_output_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_27/_input_layernorm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_27/_post_attention_layernorm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_27/_self_attention_layer/_key_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_27/_self_attention_layer/_key_norm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_27/_self_attention_layer/_output_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_27/_self_attention_layer/_query_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_27/_self_attention_layer/_query_norm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_27/_self_attention_layer/_value_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_28/_feedforward_gate_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_28/_feedforward_intermediate_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_28/_feedforward_output_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_28/_input_layernorm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_28/_linear_attn/vars": [
"model_00001.weights.h5"
],
"/layers/qwen3_5_transformer_decoder_28/_linear_attn/in_proj_a/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_28/_linear_attn/in_proj_b/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_28/_linear_attn/in_proj_qkv/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_28/_linear_attn/in_proj_z/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_28/_linear_attn/norm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_28/_linear_attn/out_proj/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_28/_post_attention_layernorm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_29/_feedforward_gate_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_29/_feedforward_intermediate_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_29/_feedforward_output_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_29/_input_layernorm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_29/_linear_attn/vars": [
"model_00001.weights.h5"
],
"/layers/qwen3_5_transformer_decoder_29/_linear_attn/in_proj_a/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_29/_linear_attn/in_proj_b/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_29/_linear_attn/in_proj_qkv/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_29/_linear_attn/in_proj_z/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_29/_linear_attn/norm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_29/_linear_attn/out_proj/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_29/_post_attention_layernorm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_30/_feedforward_gate_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_30/_feedforward_intermediate_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_30/_feedforward_output_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_30/_input_layernorm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_30/_linear_attn/vars": [
"model_00001.weights.h5"
],
"/layers/qwen3_5_transformer_decoder_30/_linear_attn/in_proj_a/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_30/_linear_attn/in_proj_b/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_30/_linear_attn/in_proj_qkv/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_30/_linear_attn/in_proj_z/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_30/_linear_attn/norm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_30/_linear_attn/out_proj/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_30/_post_attention_layernorm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_31/_feedforward_gate_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_31/_feedforward_intermediate_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_31/_feedforward_output_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_31/_input_layernorm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_31/_post_attention_layernorm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_31/_self_attention_layer/_key_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_31/_self_attention_layer/_key_norm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_31/_self_attention_layer/_output_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_31/_self_attention_layer/_query_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_31/_self_attention_layer/_query_norm/vars": "model_00001.weights.h5",
"/layers/qwen3_5_transformer_decoder_31/_self_attention_layer/_value_dense/vars": "model_00001.weights.h5",
"/layers/qwen3_5_layer_norm/vars": "model_00001.weights.h5"
}
}