deepseek_vl_1.3b_base / model.weights.json
IMvision12's picture
Upload folder using huggingface_hub
a792f44 verified
Raw
History Blame Contribute Delete
57.9 kB
{
"metadata": {
"total_size": 7011700736.0
},
"weight_map": {
"/aligner_linear1/vars": [
"deepseek_vl_1.3b_base_00000.weights.h5"
],
"/aligner_linear2/vars": [
"deepseek_vl_1.3b_base_00000.weights.h5"
],
"/decoder_layers/deepseek_vl_text_decoder_layer/attention/key/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer/attention/output_proj/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer/attention/query/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer/attention/value/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer/attention_norm/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer/mlp/down/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer/mlp/gate/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer/mlp/up/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer/mlp_norm/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_1/attention/key/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_1/attention/output_proj/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_1/attention/query/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_1/attention/value/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_1/attention_norm/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_1/mlp/down/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_1/mlp/gate/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_1/mlp/up/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_1/mlp_norm/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_2/attention/key/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_2/attention/output_proj/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_2/attention/query/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_2/attention/value/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_2/attention_norm/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_2/mlp/down/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_2/mlp/gate/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_2/mlp/up/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_2/mlp_norm/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_3/attention/key/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_3/attention/output_proj/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_3/attention/query/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_3/attention/value/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_3/attention_norm/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_3/mlp/down/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_3/mlp/gate/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_3/mlp/up/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_3/mlp_norm/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_4/attention/key/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_4/attention/output_proj/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_4/attention/query/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_4/attention/value/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_4/attention_norm/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_4/mlp/down/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_4/mlp/gate/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_4/mlp/up/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_4/mlp_norm/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_5/attention/key/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_5/attention/output_proj/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_5/attention/query/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_5/attention/value/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_5/attention_norm/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_5/mlp/down/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_5/mlp/gate/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_5/mlp/up/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_5/mlp_norm/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_6/attention/key/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_6/attention/output_proj/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_6/attention/query/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_6/attention/value/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_6/attention_norm/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_6/mlp/down/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_6/mlp/gate/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_6/mlp/up/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_6/mlp_norm/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_7/attention/key/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_7/attention/output_proj/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_7/attention/query/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_7/attention/value/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_7/attention_norm/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_7/mlp/down/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_7/mlp/gate/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_7/mlp/up/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_7/mlp_norm/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_8/attention/key/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_8/attention/output_proj/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_8/attention/query/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_8/attention/value/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_8/attention_norm/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_8/mlp/down/vars": "deepseek_vl_1.3b_base_00000.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_8/mlp/gate/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_8/mlp/up/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_8/mlp_norm/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_9/attention/key/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_9/attention/output_proj/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_9/attention/query/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_9/attention/value/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_9/attention_norm/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_9/mlp/down/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_9/mlp/gate/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_9/mlp/up/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_9/mlp_norm/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_10/attention/key/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_10/attention/output_proj/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_10/attention/query/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_10/attention/value/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_10/attention_norm/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_10/mlp/down/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_10/mlp/gate/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_10/mlp/up/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_10/mlp_norm/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_11/attention/key/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_11/attention/output_proj/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_11/attention/query/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_11/attention/value/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_11/attention_norm/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_11/mlp/down/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_11/mlp/gate/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_11/mlp/up/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_11/mlp_norm/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_12/attention/key/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_12/attention/output_proj/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_12/attention/query/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_12/attention/value/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_12/attention_norm/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_12/mlp/down/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_12/mlp/gate/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_12/mlp/up/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_12/mlp_norm/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_13/attention/key/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_13/attention/output_proj/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_13/attention/query/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_13/attention/value/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_13/attention_norm/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_13/mlp/down/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_13/mlp/gate/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_13/mlp/up/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_13/mlp_norm/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_14/attention/key/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_14/attention/output_proj/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_14/attention/query/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_14/attention/value/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_14/attention_norm/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_14/mlp/down/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_14/mlp/gate/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_14/mlp/up/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_14/mlp_norm/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_15/attention/key/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_15/attention/output_proj/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_15/attention/query/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_15/attention/value/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_15/attention_norm/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_15/mlp/down/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_15/mlp/gate/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_15/mlp/up/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_15/mlp_norm/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_16/attention/key/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_16/attention/output_proj/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_16/attention/query/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_16/attention/value/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_16/attention_norm/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_16/mlp/down/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_16/mlp/gate/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_16/mlp/up/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_16/mlp_norm/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_17/attention/key/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_17/attention/output_proj/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_17/attention/query/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_17/attention/value/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_17/attention_norm/vars": "deepseek_vl_1.3b_base_00001.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_17/mlp/down/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_17/mlp/gate/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_17/mlp/up/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_17/mlp_norm/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_18/attention/key/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_18/attention/output_proj/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_18/attention/query/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_18/attention/value/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_18/attention_norm/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_18/mlp/down/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_18/mlp/gate/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_18/mlp/up/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_18/mlp_norm/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_19/attention/key/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_19/attention/output_proj/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_19/attention/query/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_19/attention/value/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_19/attention_norm/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_19/mlp/down/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_19/mlp/gate/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_19/mlp/up/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_19/mlp_norm/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_20/attention/key/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_20/attention/output_proj/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_20/attention/query/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_20/attention/value/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_20/attention_norm/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_20/mlp/down/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_20/mlp/gate/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_20/mlp/up/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_20/mlp_norm/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_21/attention/key/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_21/attention/output_proj/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_21/attention/query/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_21/attention/value/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_21/attention_norm/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_21/mlp/down/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_21/mlp/gate/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_21/mlp/up/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_21/mlp_norm/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_22/attention/key/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_22/attention/output_proj/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_22/attention/query/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_22/attention/value/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_22/attention_norm/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_22/mlp/down/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_22/mlp/gate/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_22/mlp/up/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_22/mlp_norm/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_23/attention/key/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_23/attention/output_proj/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_23/attention/query/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_23/attention/value/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_23/attention_norm/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_23/mlp/down/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_23/mlp/gate/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_23/mlp/up/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/decoder_layers/deepseek_vl_text_decoder_layer_23/mlp_norm/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/final_norm/vars": "deepseek_vl_1.3b_base_00002.weights.h5",
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer/attention/key/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer/attention/output_proj/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer/attention/query/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer/attention/value/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer/fc1/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer/fc2/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer/layer_norm1/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer/layer_norm2/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_1/attention/key/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_1/attention/output_proj/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_1/attention/query/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_1/attention/value/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_1/fc1/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_1/fc2/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_1/layer_norm1/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_1/layer_norm2/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_2/attention/key/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_2/attention/output_proj/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_2/attention/query/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_2/attention/value/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_2/fc1/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_2/fc2/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_2/layer_norm1/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_2/layer_norm2/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_3/attention/key/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_3/attention/output_proj/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_3/attention/query/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_3/attention/value/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_3/fc1/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_3/fc2/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_3/layer_norm1/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_3/layer_norm2/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_4/attention/key/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_4/attention/output_proj/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_4/attention/query/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_4/attention/value/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_4/fc1/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_4/fc2/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_4/layer_norm1/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_4/layer_norm2/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_5/attention/key/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_5/attention/output_proj/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_5/attention/query/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_5/attention/value/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_5/fc1/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_5/fc2/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_5/layer_norm1/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_5/layer_norm2/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_6/attention/key/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_6/attention/output_proj/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_6/attention/query/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_6/attention/value/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_6/fc1/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_6/fc2/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_6/layer_norm1/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_6/layer_norm2/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_7/attention/key/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_7/attention/output_proj/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_7/attention/query/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_7/attention/value/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_7/fc1/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_7/fc2/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_7/layer_norm1/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_7/layer_norm2/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_8/attention/key/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_8/attention/output_proj/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_8/attention/query/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_8/attention/value/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_8/fc1/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_8/fc2/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_8/layer_norm1/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_8/layer_norm2/vars": [
"deepseek_vl_1.3b_base_00002.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_9/attention/key/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_9/attention/output_proj/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_9/attention/query/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_9/attention/value/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_9/fc1/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_9/fc2/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_9/layer_norm1/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_9/layer_norm2/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_10/attention/key/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_10/attention/output_proj/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_10/attention/query/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_10/attention/value/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_10/fc1/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_10/fc2/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_10/layer_norm1/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_10/layer_norm2/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_11/attention/key/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_11/attention/output_proj/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_11/attention/query/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_11/attention/value/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_11/fc1/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_11/fc2/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_11/layer_norm1/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_11/layer_norm2/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_12/attention/key/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_12/attention/output_proj/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_12/attention/query/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_12/attention/value/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_12/fc1/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_12/fc2/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_12/layer_norm1/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_12/layer_norm2/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_13/attention/key/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_13/attention/output_proj/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_13/attention/query/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_13/attention/value/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_13/fc1/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_13/fc2/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_13/layer_norm1/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_13/layer_norm2/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_14/attention/key/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_14/attention/output_proj/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_14/attention/query/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_14/attention/value/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_14/fc1/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_14/fc2/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_14/layer_norm1/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_14/layer_norm2/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_15/attention/key/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_15/attention/output_proj/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_15/attention/query/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_15/attention/value/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_15/fc1/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_15/fc2/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_15/layer_norm1/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_15/layer_norm2/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_16/attention/key/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_16/attention/output_proj/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_16/attention/query/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_16/attention/value/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_16/fc1/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_16/fc2/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_16/layer_norm1/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_16/layer_norm2/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_17/attention/key/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_17/attention/output_proj/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_17/attention/query/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_17/attention/value/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_17/fc1/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_17/fc2/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_17/layer_norm1/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_17/layer_norm2/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_18/attention/key/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_18/attention/output_proj/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_18/attention/query/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_18/attention/value/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_18/fc1/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_18/fc2/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_18/layer_norm1/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_18/layer_norm2/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_19/attention/key/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_19/attention/output_proj/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_19/attention/query/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_19/attention/value/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_19/fc1/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_19/fc2/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_19/layer_norm1/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_19/layer_norm2/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_20/attention/key/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_20/attention/output_proj/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_20/attention/query/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_20/attention/value/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_20/fc1/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_20/fc2/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_20/layer_norm1/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_20/layer_norm2/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_21/attention/key/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_21/attention/output_proj/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_21/attention/query/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_21/attention/value/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_21/fc1/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_21/fc2/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_21/layer_norm1/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_21/layer_norm2/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_22/attention/key/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_22/attention/output_proj/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_22/attention/query/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_22/attention/value/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_22/fc1/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_22/fc2/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_22/layer_norm1/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_22/layer_norm2/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_23/attention/key/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_23/attention/output_proj/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_23/attention/query/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_23/attention/value/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_23/fc1/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_23/fc2/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_23/layer_norm1/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/blocks/deepseek_vl_vision_layer_23/layer_norm2/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/patch_embed/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/deepseek_vl_vision_model/position_embedding/vars": "deepseek_vl_1.3b_base_00003.weights.h5",
"/layers/deepseek_vl_vision_model/post_layernorm/vars": [
"deepseek_vl_1.3b_base_00003.weights.h5"
],
"/layers/embedding/vars": "deepseek_vl_1.3b_base_00003.weights.h5"
}
}