Example of converting and using a Huggingface Whisper model with whisper.cpp

!git clone https://github.com/ggerganov/whisper.cpp.git

Cloning into 'whisper.cpp'...
remote: Enumerating objects: 7336, done.
remote: Total 7336 (delta 0), reused 0 (delta 0), pack-reused 7336
Receiving objects: 100% (7336/7336), 10.87 MiB | 20.88 MiB/s, done.
Resolving deltas: 100% (4753/4753), done.

%cd whisper.cpp
!make -j 8

/content/whisper.cpp
I whisper.cpp build info: 
I UNAME_S:  Linux
I UNAME_P:  x86_64
I UNAME_M:  x86_64
I CFLAGS:   -I.              -O3 -DNDEBUG -std=c11   -fPIC -D_XOPEN_SOURCE=600 -D_GNU_SOURCE -pthread -mavx -mavx2 -mfma -mf16c -msse3 -mssse3
I CXXFLAGS: -I. -I./examples -O3 -DNDEBUG -std=c++11 -fPIC -D_XOPEN_SOURCE=600 -D_GNU_SOURCE -pthread -mavx -mavx2 -mfma -mf16c -msse3 -mssse3
I LDFLAGS:  
I CC:       cc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0
I CXX:      g++ (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0

cc  -I.              -O3 -DNDEBUG -std=c11   -fPIC -D_XOPEN_SOURCE=600 -D_GNU_SOURCE -pthread -mavx -mavx2 -mfma -mf16c -msse3 -mssse3   -c ggml.c -o ggml.o
cc  -I.              -O3 -DNDEBUG -std=c11   -fPIC -D_XOPEN_SOURCE=600 -D_GNU_SOURCE -pthread -mavx -mavx2 -mfma -mf16c -msse3 -mssse3   -c ggml-alloc.c -o ggml-alloc.o
cc  -I.              -O3 -DNDEBUG -std=c11   -fPIC -D_XOPEN_SOURCE=600 -D_GNU_SOURCE -pthread -mavx -mavx2 -mfma -mf16c -msse3 -mssse3   -c ggml-backend.c -o ggml-backend.o
cc  -I.              -O3 -DNDEBUG -std=c11   -fPIC -D_XOPEN_SOURCE=600 -D_GNU_SOURCE -pthread -mavx -mavx2 -mfma -mf16c -msse3 -mssse3   -c ggml-quants.c -o ggml-quants.o
g++ -I. -I./examples -O3 -DNDEBUG -std=c++11 -fPIC -D_XOPEN_SOURCE=600 -D_GNU_SOURCE -pthread -mavx -mavx2 -mfma -mf16c -msse3 -mssse3 -c whisper.cpp -o whisper.o
g++ -I. -I./examples -O3 -DNDEBUG -std=c++11 -fPIC -D_XOPEN_SOURCE=600 -D_GNU_SOURCE -pthread -mavx -mavx2 -mfma -mf16c -msse3 -mssse3 examples/main/main.cpp examples/common.cpp examples/common-ggml.cpp ggml.o ggml-alloc.o ggml-backend.o ggml-quants.o whisper.o -o main 
g++ -I. -I./examples -O3 -DNDEBUG -std=c++11 -fPIC -D_XOPEN_SOURCE=600 -D_GNU_SOURCE -pthread -mavx -mavx2 -mfma -mf16c -msse3 -mssse3 examples/bench/bench.cpp ggml.o ggml-alloc.o ggml-backend.o ggml-quants.o whisper.o -o bench 
g++ -I. -I./examples -O3 -DNDEBUG -std=c++11 -fPIC -D_XOPEN_SOURCE=600 -D_GNU_SOURCE -pthread -mavx -mavx2 -mfma -mf16c -msse3 -mssse3 examples/quantize/quantize.cpp examples/common.cpp examples/common-ggml.cpp ggml.o ggml-alloc.o ggml-backend.o ggml-quants.o whisper.o -o quantize 
g++ -I. -I./examples -O3 -DNDEBUG -std=c++11 -fPIC -D_XOPEN_SOURCE=600 -D_GNU_SOURCE -pthread -mavx -mavx2 -mfma -mf16c -msse3 -mssse3 examples/server/server.cpp examples/common.cpp examples/common-ggml.cpp ggml.o ggml-alloc.o ggml-backend.o ggml-quants.o whisper.o -o server  
./main -h

usage: ./main [options] file0.wav file1.wav ...

options:
  -h,        --help              [default] show this help message and exit
  -t N,      --threads N         [2      ] number of threads to use during computation
  -p N,      --processors N      [1      ] number of processors to use during computation
  -ot N,     --offset-t N        [0      ] time offset in milliseconds
  -on N,     --offset-n N        [0      ] segment index offset
  -d  N,     --duration N        [0      ] duration of audio to process in milliseconds
  -mc N,     --max-context N     [-1     ] maximum number of text context tokens to store
  -ml N,     --max-len N         [0      ] maximum segment length in characters
  -sow,      --split-on-word     [false  ] split on word rather than on token
  -bo N,     --best-of N         [5      ] number of best candidates to keep
  -bs N,     --beam-size N       [5      ] beam size for beam search
  -ac N,     --audio-ctx N       [0      ] audio context size (0 - all)
  -wt N,     --word-thold N      [0.01   ] word timestamp probability threshold
  -et N,     --entropy-thold N   [2.40   ] entropy threshold for decoder fail
  -lpt N,    --logprob-thold N   [-1.00  ] log probability threshold for decoder fail
  -debug,    --debug-mode        [false  ] enable debug mode (eg. dump log_mel)
  -tr,       --translate         [false  ] translate from source language to english
  -di,       --diarize           [false  ] stereo audio diarization
  -tdrz,     --tinydiarize       [false  ] enable tinydiarize (requires a tdrz model)
  -nf,       --no-fallback       [false  ] do not use temperature fallback while decoding
  -otxt,     --output-txt        [false  ] output result in a text file
  -ovtt,     --output-vtt        [false  ] output result in a vtt file
  -osrt,     --output-srt        [false  ] output result in a srt file
  -olrc,     --output-lrc        [false  ] output result in a lrc file
  -owts,     --output-words      [false  ] output script for generating karaoke video
  -fp,       --font-path         [/System/Library/Fonts/Supplemental/Courier New Bold.ttf] path to a monospace font for karaoke video
  -ocsv,     --output-csv        [false  ] output result in a CSV file
  -oj,       --output-json       [false  ] output result in a JSON file
  -ojf,      --output-json-full  [false  ] include more information in the JSON file
  -of FNAME, --output-file FNAME [       ] output file path (without file extension)
  -np,       --no-prints         [false  ] do not print anything other than the results
  -ps,       --print-special     [false  ] print special tokens
  -pc,       --print-colors      [false  ] print colors
  -pp,       --print-progress    [false  ] print progress
  -nt,       --no-timestamps     [false  ] do not print timestamps
  -l LANG,   --language LANG     [en     ] spoken language ('auto' for auto-detect)
  -dl,       --detect-language   [false  ] exit after automatically detecting language
             --prompt PROMPT     [       ] initial prompt
  -m FNAME,  --model FNAME       [models/ggml-base.en.bin] model path
  -f FNAME,  --file FNAME        [       ] input WAV file path
  -oved D,   --ov-e-device DNAME [CPU    ] the OpenVINO device used for encode inference
  -ls,       --log-score         [false  ] log best decoder scores of tokens
  -ng,       --no-gpu            [false  ] disable GPU

!git lfs install

Updated git hooks.
Git LFS initialized.

%cd /content
!git clone https://huggingface.co/NbAiLab/whisper-large-sme

/content
Cloning into 'whisper-large-sme'...
remote: Enumerating objects: 444, done.
remote: Total 444 (delta 0), reused 0 (delta 0), pack-reused 444
Receiving objects: 100% (444/444), 642.41 KiB | 9.73 MiB/s, done.
Resolving deltas: 100% (146/146), done.

%cd /content
!git clone https://github.com/openai/whisper

/content
Cloning into 'whisper'...
remote: Enumerating objects: 712, done.
remote: Counting objects: 100% (10/10), done.
remote: Compressing objects: 100% (10/10), done.
remote: Total 712 (delta 1), reused 3 (delta 0), pack-reused 702
Receiving objects: 100% (712/712), 12.43 MiB | 20.53 MiB/s, done.
Resolving deltas: 100% (419/419), done.

!mkdir whisper-ggml-sme
!python whisper.cpp/models/convert-h5-to-ggml.py ./whisper-large-sme/ ./whisper ./whisper-ggml-sme/

mkdir: cannot create directory ‘whisper-ggml-sme’: File exists
/usr/local/lib/python3.10/dist-packages/torch/_utils.py:831: UserWarning: TypedStorage is deprecated. It will be removed in the future and UntypedStorage will be the only storage class. This should only matter to you if you are using storages directly.  To access UntypedStorage directly, use tensor.untyped_storage() instead of tensor.storage()
  return self.fget.__get__(instance, owner)()
model.encoder.conv1.weight  ->  encoder.conv1.weight
encoder.conv1.weight 3 (1280, 80, 3)
model.encoder.conv1.bias  ->  encoder.conv1.bias
  Reshaped variable:  encoder.conv1.bias  to shape:  (1280, 1)
encoder.conv1.bias 2 (1280, 1)
  Converting to float32
model.encoder.conv2.weight  ->  encoder.conv2.weight
encoder.conv2.weight 3 (1280, 1280, 3)
model.encoder.conv2.bias  ->  encoder.conv2.bias
  Reshaped variable:  encoder.conv2.bias  to shape:  (1280, 1)
encoder.conv2.bias 2 (1280, 1)
  Converting to float32
model.encoder.embed_positions.weight  ->  encoder.positional_embedding
encoder.positional_embedding 2 (1500, 1280)
  Converting to float32
model.encoder.layers.0.self_attn.k_proj.weight  ->  encoder.blocks.0.attn.key.weight
encoder.blocks.0.attn.key.weight 2 (1280, 1280)
model.encoder.layers.0.self_attn.v_proj.weight  ->  encoder.blocks.0.attn.value.weight
encoder.blocks.0.attn.value.weight 2 (1280, 1280)
model.encoder.layers.0.self_attn.v_proj.bias  ->  encoder.blocks.0.attn.value.bias
encoder.blocks.0.attn.value.bias 1 (1280,)
  Converting to float32
model.encoder.layers.0.self_attn.q_proj.weight  ->  encoder.blocks.0.attn.query.weight
encoder.blocks.0.attn.query.weight 2 (1280, 1280)
model.encoder.layers.0.self_attn.q_proj.bias  ->  encoder.blocks.0.attn.query.bias
encoder.blocks.0.attn.query.bias 1 (1280,)
  Converting to float32
model.encoder.layers.0.self_attn.out_proj.weight  ->  encoder.blocks.0.attn.out.weight
encoder.blocks.0.attn.out.weight 2 (1280, 1280)
model.encoder.layers.0.self_attn.out_proj.bias  ->  encoder.blocks.0.attn.out.bias
encoder.blocks.0.attn.out.bias 1 (1280,)
  Converting to float32
model.encoder.layers.0.self_attn_layer_norm.weight  ->  encoder.blocks.0.attn_ln.weight
encoder.blocks.0.attn_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.0.self_attn_layer_norm.bias  ->  encoder.blocks.0.attn_ln.bias
encoder.blocks.0.attn_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.0.fc1.weight  ->  encoder.blocks.0.mlp.0.weight
encoder.blocks.0.mlp.0.weight 2 (5120, 1280)
model.encoder.layers.0.fc1.bias  ->  encoder.blocks.0.mlp.0.bias
encoder.blocks.0.mlp.0.bias 1 (5120,)
  Converting to float32
model.encoder.layers.0.fc2.weight  ->  encoder.blocks.0.mlp.2.weight
encoder.blocks.0.mlp.2.weight 2 (1280, 5120)
model.encoder.layers.0.fc2.bias  ->  encoder.blocks.0.mlp.2.bias
encoder.blocks.0.mlp.2.bias 1 (1280,)
  Converting to float32
model.encoder.layers.0.final_layer_norm.weight  ->  encoder.blocks.0.mlp_ln.weight
encoder.blocks.0.mlp_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.0.final_layer_norm.bias  ->  encoder.blocks.0.mlp_ln.bias
encoder.blocks.0.mlp_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.1.self_attn.k_proj.weight  ->  encoder.blocks.1.attn.key.weight
encoder.blocks.1.attn.key.weight 2 (1280, 1280)
model.encoder.layers.1.self_attn.v_proj.weight  ->  encoder.blocks.1.attn.value.weight
encoder.blocks.1.attn.value.weight 2 (1280, 1280)
model.encoder.layers.1.self_attn.v_proj.bias  ->  encoder.blocks.1.attn.value.bias
encoder.blocks.1.attn.value.bias 1 (1280,)
  Converting to float32
model.encoder.layers.1.self_attn.q_proj.weight  ->  encoder.blocks.1.attn.query.weight
encoder.blocks.1.attn.query.weight 2 (1280, 1280)
model.encoder.layers.1.self_attn.q_proj.bias  ->  encoder.blocks.1.attn.query.bias
encoder.blocks.1.attn.query.bias 1 (1280,)
  Converting to float32
model.encoder.layers.1.self_attn.out_proj.weight  ->  encoder.blocks.1.attn.out.weight
encoder.blocks.1.attn.out.weight 2 (1280, 1280)
model.encoder.layers.1.self_attn.out_proj.bias  ->  encoder.blocks.1.attn.out.bias
encoder.blocks.1.attn.out.bias 1 (1280,)
  Converting to float32
model.encoder.layers.1.self_attn_layer_norm.weight  ->  encoder.blocks.1.attn_ln.weight
encoder.blocks.1.attn_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.1.self_attn_layer_norm.bias  ->  encoder.blocks.1.attn_ln.bias
encoder.blocks.1.attn_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.1.fc1.weight  ->  encoder.blocks.1.mlp.0.weight
encoder.blocks.1.mlp.0.weight 2 (5120, 1280)
model.encoder.layers.1.fc1.bias  ->  encoder.blocks.1.mlp.0.bias
encoder.blocks.1.mlp.0.bias 1 (5120,)
  Converting to float32
model.encoder.layers.1.fc2.weight  ->  encoder.blocks.1.mlp.2.weight
encoder.blocks.1.mlp.2.weight 2 (1280, 5120)
model.encoder.layers.1.fc2.bias  ->  encoder.blocks.1.mlp.2.bias
encoder.blocks.1.mlp.2.bias 1 (1280,)
  Converting to float32
model.encoder.layers.1.final_layer_norm.weight  ->  encoder.blocks.1.mlp_ln.weight
encoder.blocks.1.mlp_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.1.final_layer_norm.bias  ->  encoder.blocks.1.mlp_ln.bias
encoder.blocks.1.mlp_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.2.self_attn.k_proj.weight  ->  encoder.blocks.2.attn.key.weight
encoder.blocks.2.attn.key.weight 2 (1280, 1280)
model.encoder.layers.2.self_attn.v_proj.weight  ->  encoder.blocks.2.attn.value.weight
encoder.blocks.2.attn.value.weight 2 (1280, 1280)
model.encoder.layers.2.self_attn.v_proj.bias  ->  encoder.blocks.2.attn.value.bias
encoder.blocks.2.attn.value.bias 1 (1280,)
  Converting to float32
model.encoder.layers.2.self_attn.q_proj.weight  ->  encoder.blocks.2.attn.query.weight
encoder.blocks.2.attn.query.weight 2 (1280, 1280)
model.encoder.layers.2.self_attn.q_proj.bias  ->  encoder.blocks.2.attn.query.bias
encoder.blocks.2.attn.query.bias 1 (1280,)
  Converting to float32
model.encoder.layers.2.self_attn.out_proj.weight  ->  encoder.blocks.2.attn.out.weight
encoder.blocks.2.attn.out.weight 2 (1280, 1280)
model.encoder.layers.2.self_attn.out_proj.bias  ->  encoder.blocks.2.attn.out.bias
encoder.blocks.2.attn.out.bias 1 (1280,)
  Converting to float32
model.encoder.layers.2.self_attn_layer_norm.weight  ->  encoder.blocks.2.attn_ln.weight
encoder.blocks.2.attn_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.2.self_attn_layer_norm.bias  ->  encoder.blocks.2.attn_ln.bias
encoder.blocks.2.attn_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.2.fc1.weight  ->  encoder.blocks.2.mlp.0.weight
encoder.blocks.2.mlp.0.weight 2 (5120, 1280)
model.encoder.layers.2.fc1.bias  ->  encoder.blocks.2.mlp.0.bias
encoder.blocks.2.mlp.0.bias 1 (5120,)
  Converting to float32
model.encoder.layers.2.fc2.weight  ->  encoder.blocks.2.mlp.2.weight
encoder.blocks.2.mlp.2.weight 2 (1280, 5120)
model.encoder.layers.2.fc2.bias  ->  encoder.blocks.2.mlp.2.bias
encoder.blocks.2.mlp.2.bias 1 (1280,)
  Converting to float32
model.encoder.layers.2.final_layer_norm.weight  ->  encoder.blocks.2.mlp_ln.weight
encoder.blocks.2.mlp_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.2.final_layer_norm.bias  ->  encoder.blocks.2.mlp_ln.bias
encoder.blocks.2.mlp_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.3.self_attn.k_proj.weight  ->  encoder.blocks.3.attn.key.weight
encoder.blocks.3.attn.key.weight 2 (1280, 1280)
model.encoder.layers.3.self_attn.v_proj.weight  ->  encoder.blocks.3.attn.value.weight
encoder.blocks.3.attn.value.weight 2 (1280, 1280)
model.encoder.layers.3.self_attn.v_proj.bias  ->  encoder.blocks.3.attn.value.bias
encoder.blocks.3.attn.value.bias 1 (1280,)
  Converting to float32
model.encoder.layers.3.self_attn.q_proj.weight  ->  encoder.blocks.3.attn.query.weight
encoder.blocks.3.attn.query.weight 2 (1280, 1280)
model.encoder.layers.3.self_attn.q_proj.bias  ->  encoder.blocks.3.attn.query.bias
encoder.blocks.3.attn.query.bias 1 (1280,)
  Converting to float32
model.encoder.layers.3.self_attn.out_proj.weight  ->  encoder.blocks.3.attn.out.weight
encoder.blocks.3.attn.out.weight 2 (1280, 1280)
model.encoder.layers.3.self_attn.out_proj.bias  ->  encoder.blocks.3.attn.out.bias
encoder.blocks.3.attn.out.bias 1 (1280,)
  Converting to float32
model.encoder.layers.3.self_attn_layer_norm.weight  ->  encoder.blocks.3.attn_ln.weight
encoder.blocks.3.attn_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.3.self_attn_layer_norm.bias  ->  encoder.blocks.3.attn_ln.bias
encoder.blocks.3.attn_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.3.fc1.weight  ->  encoder.blocks.3.mlp.0.weight
encoder.blocks.3.mlp.0.weight 2 (5120, 1280)
model.encoder.layers.3.fc1.bias  ->  encoder.blocks.3.mlp.0.bias
encoder.blocks.3.mlp.0.bias 1 (5120,)
  Converting to float32
model.encoder.layers.3.fc2.weight  ->  encoder.blocks.3.mlp.2.weight
encoder.blocks.3.mlp.2.weight 2 (1280, 5120)
model.encoder.layers.3.fc2.bias  ->  encoder.blocks.3.mlp.2.bias
encoder.blocks.3.mlp.2.bias 1 (1280,)
  Converting to float32
model.encoder.layers.3.final_layer_norm.weight  ->  encoder.blocks.3.mlp_ln.weight
encoder.blocks.3.mlp_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.3.final_layer_norm.bias  ->  encoder.blocks.3.mlp_ln.bias
encoder.blocks.3.mlp_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.4.self_attn.k_proj.weight  ->  encoder.blocks.4.attn.key.weight
encoder.blocks.4.attn.key.weight 2 (1280, 1280)
model.encoder.layers.4.self_attn.v_proj.weight  ->  encoder.blocks.4.attn.value.weight
encoder.blocks.4.attn.value.weight 2 (1280, 1280)
model.encoder.layers.4.self_attn.v_proj.bias  ->  encoder.blocks.4.attn.value.bias
encoder.blocks.4.attn.value.bias 1 (1280,)
  Converting to float32
model.encoder.layers.4.self_attn.q_proj.weight  ->  encoder.blocks.4.attn.query.weight
encoder.blocks.4.attn.query.weight 2 (1280, 1280)
model.encoder.layers.4.self_attn.q_proj.bias  ->  encoder.blocks.4.attn.query.bias
encoder.blocks.4.attn.query.bias 1 (1280,)
  Converting to float32
model.encoder.layers.4.self_attn.out_proj.weight  ->  encoder.blocks.4.attn.out.weight
encoder.blocks.4.attn.out.weight 2 (1280, 1280)
model.encoder.layers.4.self_attn.out_proj.bias  ->  encoder.blocks.4.attn.out.bias
encoder.blocks.4.attn.out.bias 1 (1280,)
  Converting to float32
model.encoder.layers.4.self_attn_layer_norm.weight  ->  encoder.blocks.4.attn_ln.weight
encoder.blocks.4.attn_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.4.self_attn_layer_norm.bias  ->  encoder.blocks.4.attn_ln.bias
encoder.blocks.4.attn_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.4.fc1.weight  ->  encoder.blocks.4.mlp.0.weight
encoder.blocks.4.mlp.0.weight 2 (5120, 1280)
model.encoder.layers.4.fc1.bias  ->  encoder.blocks.4.mlp.0.bias
encoder.blocks.4.mlp.0.bias 1 (5120,)
  Converting to float32
model.encoder.layers.4.fc2.weight  ->  encoder.blocks.4.mlp.2.weight
encoder.blocks.4.mlp.2.weight 2 (1280, 5120)
model.encoder.layers.4.fc2.bias  ->  encoder.blocks.4.mlp.2.bias
encoder.blocks.4.mlp.2.bias 1 (1280,)
  Converting to float32
model.encoder.layers.4.final_layer_norm.weight  ->  encoder.blocks.4.mlp_ln.weight
encoder.blocks.4.mlp_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.4.final_layer_norm.bias  ->  encoder.blocks.4.mlp_ln.bias
encoder.blocks.4.mlp_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.5.self_attn.k_proj.weight  ->  encoder.blocks.5.attn.key.weight
encoder.blocks.5.attn.key.weight 2 (1280, 1280)
model.encoder.layers.5.self_attn.v_proj.weight  ->  encoder.blocks.5.attn.value.weight
encoder.blocks.5.attn.value.weight 2 (1280, 1280)
model.encoder.layers.5.self_attn.v_proj.bias  ->  encoder.blocks.5.attn.value.bias
encoder.blocks.5.attn.value.bias 1 (1280,)
  Converting to float32
model.encoder.layers.5.self_attn.q_proj.weight  ->  encoder.blocks.5.attn.query.weight
encoder.blocks.5.attn.query.weight 2 (1280, 1280)
model.encoder.layers.5.self_attn.q_proj.bias  ->  encoder.blocks.5.attn.query.bias
encoder.blocks.5.attn.query.bias 1 (1280,)
  Converting to float32
model.encoder.layers.5.self_attn.out_proj.weight  ->  encoder.blocks.5.attn.out.weight
encoder.blocks.5.attn.out.weight 2 (1280, 1280)
model.encoder.layers.5.self_attn.out_proj.bias  ->  encoder.blocks.5.attn.out.bias
encoder.blocks.5.attn.out.bias 1 (1280,)
  Converting to float32
model.encoder.layers.5.self_attn_layer_norm.weight  ->  encoder.blocks.5.attn_ln.weight
encoder.blocks.5.attn_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.5.self_attn_layer_norm.bias  ->  encoder.blocks.5.attn_ln.bias
encoder.blocks.5.attn_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.5.fc1.weight  ->  encoder.blocks.5.mlp.0.weight
encoder.blocks.5.mlp.0.weight 2 (5120, 1280)
model.encoder.layers.5.fc1.bias  ->  encoder.blocks.5.mlp.0.bias
encoder.blocks.5.mlp.0.bias 1 (5120,)
  Converting to float32
model.encoder.layers.5.fc2.weight  ->  encoder.blocks.5.mlp.2.weight
encoder.blocks.5.mlp.2.weight 2 (1280, 5120)
model.encoder.layers.5.fc2.bias  ->  encoder.blocks.5.mlp.2.bias
encoder.blocks.5.mlp.2.bias 1 (1280,)
  Converting to float32
model.encoder.layers.5.final_layer_norm.weight  ->  encoder.blocks.5.mlp_ln.weight
encoder.blocks.5.mlp_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.5.final_layer_norm.bias  ->  encoder.blocks.5.mlp_ln.bias
encoder.blocks.5.mlp_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.6.self_attn.k_proj.weight  ->  encoder.blocks.6.attn.key.weight
encoder.blocks.6.attn.key.weight 2 (1280, 1280)
model.encoder.layers.6.self_attn.v_proj.weight  ->  encoder.blocks.6.attn.value.weight
encoder.blocks.6.attn.value.weight 2 (1280, 1280)
model.encoder.layers.6.self_attn.v_proj.bias  ->  encoder.blocks.6.attn.value.bias
encoder.blocks.6.attn.value.bias 1 (1280,)
  Converting to float32
model.encoder.layers.6.self_attn.q_proj.weight  ->  encoder.blocks.6.attn.query.weight
encoder.blocks.6.attn.query.weight 2 (1280, 1280)
model.encoder.layers.6.self_attn.q_proj.bias  ->  encoder.blocks.6.attn.query.bias
encoder.blocks.6.attn.query.bias 1 (1280,)
  Converting to float32
model.encoder.layers.6.self_attn.out_proj.weight  ->  encoder.blocks.6.attn.out.weight
encoder.blocks.6.attn.out.weight 2 (1280, 1280)
model.encoder.layers.6.self_attn.out_proj.bias  ->  encoder.blocks.6.attn.out.bias
encoder.blocks.6.attn.out.bias 1 (1280,)
  Converting to float32
model.encoder.layers.6.self_attn_layer_norm.weight  ->  encoder.blocks.6.attn_ln.weight
encoder.blocks.6.attn_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.6.self_attn_layer_norm.bias  ->  encoder.blocks.6.attn_ln.bias
encoder.blocks.6.attn_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.6.fc1.weight  ->  encoder.blocks.6.mlp.0.weight
encoder.blocks.6.mlp.0.weight 2 (5120, 1280)
model.encoder.layers.6.fc1.bias  ->  encoder.blocks.6.mlp.0.bias
encoder.blocks.6.mlp.0.bias 1 (5120,)
  Converting to float32
model.encoder.layers.6.fc2.weight  ->  encoder.blocks.6.mlp.2.weight
encoder.blocks.6.mlp.2.weight 2 (1280, 5120)
model.encoder.layers.6.fc2.bias  ->  encoder.blocks.6.mlp.2.bias
encoder.blocks.6.mlp.2.bias 1 (1280,)
  Converting to float32
model.encoder.layers.6.final_layer_norm.weight  ->  encoder.blocks.6.mlp_ln.weight
encoder.blocks.6.mlp_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.6.final_layer_norm.bias  ->  encoder.blocks.6.mlp_ln.bias
encoder.blocks.6.mlp_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.7.self_attn.k_proj.weight  ->  encoder.blocks.7.attn.key.weight
encoder.blocks.7.attn.key.weight 2 (1280, 1280)
model.encoder.layers.7.self_attn.v_proj.weight  ->  encoder.blocks.7.attn.value.weight
encoder.blocks.7.attn.value.weight 2 (1280, 1280)
model.encoder.layers.7.self_attn.v_proj.bias  ->  encoder.blocks.7.attn.value.bias
encoder.blocks.7.attn.value.bias 1 (1280,)
  Converting to float32
model.encoder.layers.7.self_attn.q_proj.weight  ->  encoder.blocks.7.attn.query.weight
encoder.blocks.7.attn.query.weight 2 (1280, 1280)
model.encoder.layers.7.self_attn.q_proj.bias  ->  encoder.blocks.7.attn.query.bias
encoder.blocks.7.attn.query.bias 1 (1280,)
  Converting to float32
model.encoder.layers.7.self_attn.out_proj.weight  ->  encoder.blocks.7.attn.out.weight
encoder.blocks.7.attn.out.weight 2 (1280, 1280)
model.encoder.layers.7.self_attn.out_proj.bias  ->  encoder.blocks.7.attn.out.bias
encoder.blocks.7.attn.out.bias 1 (1280,)
  Converting to float32
model.encoder.layers.7.self_attn_layer_norm.weight  ->  encoder.blocks.7.attn_ln.weight
encoder.blocks.7.attn_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.7.self_attn_layer_norm.bias  ->  encoder.blocks.7.attn_ln.bias
encoder.blocks.7.attn_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.7.fc1.weight  ->  encoder.blocks.7.mlp.0.weight
encoder.blocks.7.mlp.0.weight 2 (5120, 1280)
model.encoder.layers.7.fc1.bias  ->  encoder.blocks.7.mlp.0.bias
encoder.blocks.7.mlp.0.bias 1 (5120,)
  Converting to float32
model.encoder.layers.7.fc2.weight  ->  encoder.blocks.7.mlp.2.weight
encoder.blocks.7.mlp.2.weight 2 (1280, 5120)
model.encoder.layers.7.fc2.bias  ->  encoder.blocks.7.mlp.2.bias
encoder.blocks.7.mlp.2.bias 1 (1280,)
  Converting to float32
model.encoder.layers.7.final_layer_norm.weight  ->  encoder.blocks.7.mlp_ln.weight
encoder.blocks.7.mlp_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.7.final_layer_norm.bias  ->  encoder.blocks.7.mlp_ln.bias
encoder.blocks.7.mlp_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.8.self_attn.k_proj.weight  ->  encoder.blocks.8.attn.key.weight
encoder.blocks.8.attn.key.weight 2 (1280, 1280)
model.encoder.layers.8.self_attn.v_proj.weight  ->  encoder.blocks.8.attn.value.weight
encoder.blocks.8.attn.value.weight 2 (1280, 1280)
model.encoder.layers.8.self_attn.v_proj.bias  ->  encoder.blocks.8.attn.value.bias
encoder.blocks.8.attn.value.bias 1 (1280,)
  Converting to float32
model.encoder.layers.8.self_attn.q_proj.weight  ->  encoder.blocks.8.attn.query.weight
encoder.blocks.8.attn.query.weight 2 (1280, 1280)
model.encoder.layers.8.self_attn.q_proj.bias  ->  encoder.blocks.8.attn.query.bias
encoder.blocks.8.attn.query.bias 1 (1280,)
  Converting to float32
model.encoder.layers.8.self_attn.out_proj.weight  ->  encoder.blocks.8.attn.out.weight
encoder.blocks.8.attn.out.weight 2 (1280, 1280)
model.encoder.layers.8.self_attn.out_proj.bias  ->  encoder.blocks.8.attn.out.bias
encoder.blocks.8.attn.out.bias 1 (1280,)
  Converting to float32
model.encoder.layers.8.self_attn_layer_norm.weight  ->  encoder.blocks.8.attn_ln.weight
encoder.blocks.8.attn_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.8.self_attn_layer_norm.bias  ->  encoder.blocks.8.attn_ln.bias
encoder.blocks.8.attn_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.8.fc1.weight  ->  encoder.blocks.8.mlp.0.weight
encoder.blocks.8.mlp.0.weight 2 (5120, 1280)
model.encoder.layers.8.fc1.bias  ->  encoder.blocks.8.mlp.0.bias
encoder.blocks.8.mlp.0.bias 1 (5120,)
  Converting to float32
model.encoder.layers.8.fc2.weight  ->  encoder.blocks.8.mlp.2.weight
encoder.blocks.8.mlp.2.weight 2 (1280, 5120)
model.encoder.layers.8.fc2.bias  ->  encoder.blocks.8.mlp.2.bias
encoder.blocks.8.mlp.2.bias 1 (1280,)
  Converting to float32
model.encoder.layers.8.final_layer_norm.weight  ->  encoder.blocks.8.mlp_ln.weight
encoder.blocks.8.mlp_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.8.final_layer_norm.bias  ->  encoder.blocks.8.mlp_ln.bias
encoder.blocks.8.mlp_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.9.self_attn.k_proj.weight  ->  encoder.blocks.9.attn.key.weight
encoder.blocks.9.attn.key.weight 2 (1280, 1280)
model.encoder.layers.9.self_attn.v_proj.weight  ->  encoder.blocks.9.attn.value.weight
encoder.blocks.9.attn.value.weight 2 (1280, 1280)
model.encoder.layers.9.self_attn.v_proj.bias  ->  encoder.blocks.9.attn.value.bias
encoder.blocks.9.attn.value.bias 1 (1280,)
  Converting to float32
model.encoder.layers.9.self_attn.q_proj.weight  ->  encoder.blocks.9.attn.query.weight
encoder.blocks.9.attn.query.weight 2 (1280, 1280)
model.encoder.layers.9.self_attn.q_proj.bias  ->  encoder.blocks.9.attn.query.bias
encoder.blocks.9.attn.query.bias 1 (1280,)
  Converting to float32
model.encoder.layers.9.self_attn.out_proj.weight  ->  encoder.blocks.9.attn.out.weight
encoder.blocks.9.attn.out.weight 2 (1280, 1280)
model.encoder.layers.9.self_attn.out_proj.bias  ->  encoder.blocks.9.attn.out.bias
encoder.blocks.9.attn.out.bias 1 (1280,)
  Converting to float32
model.encoder.layers.9.self_attn_layer_norm.weight  ->  encoder.blocks.9.attn_ln.weight
encoder.blocks.9.attn_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.9.self_attn_layer_norm.bias  ->  encoder.blocks.9.attn_ln.bias
encoder.blocks.9.attn_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.9.fc1.weight  ->  encoder.blocks.9.mlp.0.weight
encoder.blocks.9.mlp.0.weight 2 (5120, 1280)
model.encoder.layers.9.fc1.bias  ->  encoder.blocks.9.mlp.0.bias
encoder.blocks.9.mlp.0.bias 1 (5120,)
  Converting to float32
model.encoder.layers.9.fc2.weight  ->  encoder.blocks.9.mlp.2.weight
encoder.blocks.9.mlp.2.weight 2 (1280, 5120)
model.encoder.layers.9.fc2.bias  ->  encoder.blocks.9.mlp.2.bias
encoder.blocks.9.mlp.2.bias 1 (1280,)
  Converting to float32
model.encoder.layers.9.final_layer_norm.weight  ->  encoder.blocks.9.mlp_ln.weight
encoder.blocks.9.mlp_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.9.final_layer_norm.bias  ->  encoder.blocks.9.mlp_ln.bias
encoder.blocks.9.mlp_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.10.self_attn.k_proj.weight  ->  encoder.blocks.10.attn.key.weight
encoder.blocks.10.attn.key.weight 2 (1280, 1280)
model.encoder.layers.10.self_attn.v_proj.weight  ->  encoder.blocks.10.attn.value.weight
encoder.blocks.10.attn.value.weight 2 (1280, 1280)
model.encoder.layers.10.self_attn.v_proj.bias  ->  encoder.blocks.10.attn.value.bias
encoder.blocks.10.attn.value.bias 1 (1280,)
  Converting to float32
model.encoder.layers.10.self_attn.q_proj.weight  ->  encoder.blocks.10.attn.query.weight
encoder.blocks.10.attn.query.weight 2 (1280, 1280)
model.encoder.layers.10.self_attn.q_proj.bias  ->  encoder.blocks.10.attn.query.bias
encoder.blocks.10.attn.query.bias 1 (1280,)
  Converting to float32
model.encoder.layers.10.self_attn.out_proj.weight  ->  encoder.blocks.10.attn.out.weight
encoder.blocks.10.attn.out.weight 2 (1280, 1280)
model.encoder.layers.10.self_attn.out_proj.bias  ->  encoder.blocks.10.attn.out.bias
encoder.blocks.10.attn.out.bias 1 (1280,)
  Converting to float32
model.encoder.layers.10.self_attn_layer_norm.weight  ->  encoder.blocks.10.attn_ln.weight
encoder.blocks.10.attn_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.10.self_attn_layer_norm.bias  ->  encoder.blocks.10.attn_ln.bias
encoder.blocks.10.attn_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.10.fc1.weight  ->  encoder.blocks.10.mlp.0.weight
encoder.blocks.10.mlp.0.weight 2 (5120, 1280)
model.encoder.layers.10.fc1.bias  ->  encoder.blocks.10.mlp.0.bias
encoder.blocks.10.mlp.0.bias 1 (5120,)
  Converting to float32
model.encoder.layers.10.fc2.weight  ->  encoder.blocks.10.mlp.2.weight
encoder.blocks.10.mlp.2.weight 2 (1280, 5120)
model.encoder.layers.10.fc2.bias  ->  encoder.blocks.10.mlp.2.bias
encoder.blocks.10.mlp.2.bias 1 (1280,)
  Converting to float32
model.encoder.layers.10.final_layer_norm.weight  ->  encoder.blocks.10.mlp_ln.weight
encoder.blocks.10.mlp_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.10.final_layer_norm.bias  ->  encoder.blocks.10.mlp_ln.bias
encoder.blocks.10.mlp_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.11.self_attn.k_proj.weight  ->  encoder.blocks.11.attn.key.weight
encoder.blocks.11.attn.key.weight 2 (1280, 1280)
model.encoder.layers.11.self_attn.v_proj.weight  ->  encoder.blocks.11.attn.value.weight
encoder.blocks.11.attn.value.weight 2 (1280, 1280)
model.encoder.layers.11.self_attn.v_proj.bias  ->  encoder.blocks.11.attn.value.bias
encoder.blocks.11.attn.value.bias 1 (1280,)
  Converting to float32
model.encoder.layers.11.self_attn.q_proj.weight  ->  encoder.blocks.11.attn.query.weight
encoder.blocks.11.attn.query.weight 2 (1280, 1280)
model.encoder.layers.11.self_attn.q_proj.bias  ->  encoder.blocks.11.attn.query.bias
encoder.blocks.11.attn.query.bias 1 (1280,)
  Converting to float32
model.encoder.layers.11.self_attn.out_proj.weight  ->  encoder.blocks.11.attn.out.weight
encoder.blocks.11.attn.out.weight 2 (1280, 1280)
model.encoder.layers.11.self_attn.out_proj.bias  ->  encoder.blocks.11.attn.out.bias
encoder.blocks.11.attn.out.bias 1 (1280,)
  Converting to float32
model.encoder.layers.11.self_attn_layer_norm.weight  ->  encoder.blocks.11.attn_ln.weight
encoder.blocks.11.attn_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.11.self_attn_layer_norm.bias  ->  encoder.blocks.11.attn_ln.bias
encoder.blocks.11.attn_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.11.fc1.weight  ->  encoder.blocks.11.mlp.0.weight
encoder.blocks.11.mlp.0.weight 2 (5120, 1280)
model.encoder.layers.11.fc1.bias  ->  encoder.blocks.11.mlp.0.bias
encoder.blocks.11.mlp.0.bias 1 (5120,)
  Converting to float32
model.encoder.layers.11.fc2.weight  ->  encoder.blocks.11.mlp.2.weight
encoder.blocks.11.mlp.2.weight 2 (1280, 5120)
model.encoder.layers.11.fc2.bias  ->  encoder.blocks.11.mlp.2.bias
encoder.blocks.11.mlp.2.bias 1 (1280,)
  Converting to float32
model.encoder.layers.11.final_layer_norm.weight  ->  encoder.blocks.11.mlp_ln.weight
encoder.blocks.11.mlp_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.11.final_layer_norm.bias  ->  encoder.blocks.11.mlp_ln.bias
encoder.blocks.11.mlp_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.12.self_attn.k_proj.weight  ->  encoder.blocks.12.attn.key.weight
encoder.blocks.12.attn.key.weight 2 (1280, 1280)
model.encoder.layers.12.self_attn.v_proj.weight  ->  encoder.blocks.12.attn.value.weight
encoder.blocks.12.attn.value.weight 2 (1280, 1280)
model.encoder.layers.12.self_attn.v_proj.bias  ->  encoder.blocks.12.attn.value.bias
encoder.blocks.12.attn.value.bias 1 (1280,)
  Converting to float32
model.encoder.layers.12.self_attn.q_proj.weight  ->  encoder.blocks.12.attn.query.weight
encoder.blocks.12.attn.query.weight 2 (1280, 1280)
model.encoder.layers.12.self_attn.q_proj.bias  ->  encoder.blocks.12.attn.query.bias
encoder.blocks.12.attn.query.bias 1 (1280,)
  Converting to float32
model.encoder.layers.12.self_attn.out_proj.weight  ->  encoder.blocks.12.attn.out.weight
encoder.blocks.12.attn.out.weight 2 (1280, 1280)
model.encoder.layers.12.self_attn.out_proj.bias  ->  encoder.blocks.12.attn.out.bias
encoder.blocks.12.attn.out.bias 1 (1280,)
  Converting to float32
model.encoder.layers.12.self_attn_layer_norm.weight  ->  encoder.blocks.12.attn_ln.weight
encoder.blocks.12.attn_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.12.self_attn_layer_norm.bias  ->  encoder.blocks.12.attn_ln.bias
encoder.blocks.12.attn_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.12.fc1.weight  ->  encoder.blocks.12.mlp.0.weight
encoder.blocks.12.mlp.0.weight 2 (5120, 1280)
model.encoder.layers.12.fc1.bias  ->  encoder.blocks.12.mlp.0.bias
encoder.blocks.12.mlp.0.bias 1 (5120,)
  Converting to float32
model.encoder.layers.12.fc2.weight  ->  encoder.blocks.12.mlp.2.weight
encoder.blocks.12.mlp.2.weight 2 (1280, 5120)
model.encoder.layers.12.fc2.bias  ->  encoder.blocks.12.mlp.2.bias
encoder.blocks.12.mlp.2.bias 1 (1280,)
  Converting to float32
model.encoder.layers.12.final_layer_norm.weight  ->  encoder.blocks.12.mlp_ln.weight
encoder.blocks.12.mlp_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.12.final_layer_norm.bias  ->  encoder.blocks.12.mlp_ln.bias
encoder.blocks.12.mlp_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.13.self_attn.k_proj.weight  ->  encoder.blocks.13.attn.key.weight
encoder.blocks.13.attn.key.weight 2 (1280, 1280)
model.encoder.layers.13.self_attn.v_proj.weight  ->  encoder.blocks.13.attn.value.weight
encoder.blocks.13.attn.value.weight 2 (1280, 1280)
model.encoder.layers.13.self_attn.v_proj.bias  ->  encoder.blocks.13.attn.value.bias
encoder.blocks.13.attn.value.bias 1 (1280,)
  Converting to float32
model.encoder.layers.13.self_attn.q_proj.weight  ->  encoder.blocks.13.attn.query.weight
encoder.blocks.13.attn.query.weight 2 (1280, 1280)
model.encoder.layers.13.self_attn.q_proj.bias  ->  encoder.blocks.13.attn.query.bias
encoder.blocks.13.attn.query.bias 1 (1280,)
  Converting to float32
model.encoder.layers.13.self_attn.out_proj.weight  ->  encoder.blocks.13.attn.out.weight
encoder.blocks.13.attn.out.weight 2 (1280, 1280)
model.encoder.layers.13.self_attn.out_proj.bias  ->  encoder.blocks.13.attn.out.bias
encoder.blocks.13.attn.out.bias 1 (1280,)
  Converting to float32
model.encoder.layers.13.self_attn_layer_norm.weight  ->  encoder.blocks.13.attn_ln.weight
encoder.blocks.13.attn_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.13.self_attn_layer_norm.bias  ->  encoder.blocks.13.attn_ln.bias
encoder.blocks.13.attn_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.13.fc1.weight  ->  encoder.blocks.13.mlp.0.weight
encoder.blocks.13.mlp.0.weight 2 (5120, 1280)
model.encoder.layers.13.fc1.bias  ->  encoder.blocks.13.mlp.0.bias
encoder.blocks.13.mlp.0.bias 1 (5120,)
  Converting to float32
model.encoder.layers.13.fc2.weight  ->  encoder.blocks.13.mlp.2.weight
encoder.blocks.13.mlp.2.weight 2 (1280, 5120)
model.encoder.layers.13.fc2.bias  ->  encoder.blocks.13.mlp.2.bias
encoder.blocks.13.mlp.2.bias 1 (1280,)
  Converting to float32
model.encoder.layers.13.final_layer_norm.weight  ->  encoder.blocks.13.mlp_ln.weight
encoder.blocks.13.mlp_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.13.final_layer_norm.bias  ->  encoder.blocks.13.mlp_ln.bias
encoder.blocks.13.mlp_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.14.self_attn.k_proj.weight  ->  encoder.blocks.14.attn.key.weight
encoder.blocks.14.attn.key.weight 2 (1280, 1280)
model.encoder.layers.14.self_attn.v_proj.weight  ->  encoder.blocks.14.attn.value.weight
encoder.blocks.14.attn.value.weight 2 (1280, 1280)
model.encoder.layers.14.self_attn.v_proj.bias  ->  encoder.blocks.14.attn.value.bias
encoder.blocks.14.attn.value.bias 1 (1280,)
  Converting to float32
model.encoder.layers.14.self_attn.q_proj.weight  ->  encoder.blocks.14.attn.query.weight
encoder.blocks.14.attn.query.weight 2 (1280, 1280)
model.encoder.layers.14.self_attn.q_proj.bias  ->  encoder.blocks.14.attn.query.bias
encoder.blocks.14.attn.query.bias 1 (1280,)
  Converting to float32
model.encoder.layers.14.self_attn.out_proj.weight  ->  encoder.blocks.14.attn.out.weight
encoder.blocks.14.attn.out.weight 2 (1280, 1280)
model.encoder.layers.14.self_attn.out_proj.bias  ->  encoder.blocks.14.attn.out.bias
encoder.blocks.14.attn.out.bias 1 (1280,)
  Converting to float32
model.encoder.layers.14.self_attn_layer_norm.weight  ->  encoder.blocks.14.attn_ln.weight
encoder.blocks.14.attn_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.14.self_attn_layer_norm.bias  ->  encoder.blocks.14.attn_ln.bias
encoder.blocks.14.attn_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.14.fc1.weight  ->  encoder.blocks.14.mlp.0.weight
encoder.blocks.14.mlp.0.weight 2 (5120, 1280)
model.encoder.layers.14.fc1.bias  ->  encoder.blocks.14.mlp.0.bias
encoder.blocks.14.mlp.0.bias 1 (5120,)
  Converting to float32
model.encoder.layers.14.fc2.weight  ->  encoder.blocks.14.mlp.2.weight
encoder.blocks.14.mlp.2.weight 2 (1280, 5120)
model.encoder.layers.14.fc2.bias  ->  encoder.blocks.14.mlp.2.bias
encoder.blocks.14.mlp.2.bias 1 (1280,)
  Converting to float32
model.encoder.layers.14.final_layer_norm.weight  ->  encoder.blocks.14.mlp_ln.weight
encoder.blocks.14.mlp_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.14.final_layer_norm.bias  ->  encoder.blocks.14.mlp_ln.bias
encoder.blocks.14.mlp_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.15.self_attn.k_proj.weight  ->  encoder.blocks.15.attn.key.weight
encoder.blocks.15.attn.key.weight 2 (1280, 1280)
model.encoder.layers.15.self_attn.v_proj.weight  ->  encoder.blocks.15.attn.value.weight
encoder.blocks.15.attn.value.weight 2 (1280, 1280)
model.encoder.layers.15.self_attn.v_proj.bias  ->  encoder.blocks.15.attn.value.bias
encoder.blocks.15.attn.value.bias 1 (1280,)
  Converting to float32
model.encoder.layers.15.self_attn.q_proj.weight  ->  encoder.blocks.15.attn.query.weight
encoder.blocks.15.attn.query.weight 2 (1280, 1280)
model.encoder.layers.15.self_attn.q_proj.bias  ->  encoder.blocks.15.attn.query.bias
encoder.blocks.15.attn.query.bias 1 (1280,)
  Converting to float32
model.encoder.layers.15.self_attn.out_proj.weight  ->  encoder.blocks.15.attn.out.weight
encoder.blocks.15.attn.out.weight 2 (1280, 1280)
model.encoder.layers.15.self_attn.out_proj.bias  ->  encoder.blocks.15.attn.out.bias
encoder.blocks.15.attn.out.bias 1 (1280,)
  Converting to float32
model.encoder.layers.15.self_attn_layer_norm.weight  ->  encoder.blocks.15.attn_ln.weight
encoder.blocks.15.attn_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.15.self_attn_layer_norm.bias  ->  encoder.blocks.15.attn_ln.bias
encoder.blocks.15.attn_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.15.fc1.weight  ->  encoder.blocks.15.mlp.0.weight
encoder.blocks.15.mlp.0.weight 2 (5120, 1280)
model.encoder.layers.15.fc1.bias  ->  encoder.blocks.15.mlp.0.bias
encoder.blocks.15.mlp.0.bias 1 (5120,)
  Converting to float32
model.encoder.layers.15.fc2.weight  ->  encoder.blocks.15.mlp.2.weight
encoder.blocks.15.mlp.2.weight 2 (1280, 5120)
model.encoder.layers.15.fc2.bias  ->  encoder.blocks.15.mlp.2.bias
encoder.blocks.15.mlp.2.bias 1 (1280,)
  Converting to float32
model.encoder.layers.15.final_layer_norm.weight  ->  encoder.blocks.15.mlp_ln.weight
encoder.blocks.15.mlp_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.15.final_layer_norm.bias  ->  encoder.blocks.15.mlp_ln.bias
encoder.blocks.15.mlp_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.16.self_attn.k_proj.weight  ->  encoder.blocks.16.attn.key.weight
encoder.blocks.16.attn.key.weight 2 (1280, 1280)
model.encoder.layers.16.self_attn.v_proj.weight  ->  encoder.blocks.16.attn.value.weight
encoder.blocks.16.attn.value.weight 2 (1280, 1280)
model.encoder.layers.16.self_attn.v_proj.bias  ->  encoder.blocks.16.attn.value.bias
encoder.blocks.16.attn.value.bias 1 (1280,)
  Converting to float32
model.encoder.layers.16.self_attn.q_proj.weight  ->  encoder.blocks.16.attn.query.weight
encoder.blocks.16.attn.query.weight 2 (1280, 1280)
model.encoder.layers.16.self_attn.q_proj.bias  ->  encoder.blocks.16.attn.query.bias
encoder.blocks.16.attn.query.bias 1 (1280,)
  Converting to float32
model.encoder.layers.16.self_attn.out_proj.weight  ->  encoder.blocks.16.attn.out.weight
encoder.blocks.16.attn.out.weight 2 (1280, 1280)
model.encoder.layers.16.self_attn.out_proj.bias  ->  encoder.blocks.16.attn.out.bias
encoder.blocks.16.attn.out.bias 1 (1280,)
  Converting to float32
model.encoder.layers.16.self_attn_layer_norm.weight  ->  encoder.blocks.16.attn_ln.weight
encoder.blocks.16.attn_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.16.self_attn_layer_norm.bias  ->  encoder.blocks.16.attn_ln.bias
encoder.blocks.16.attn_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.16.fc1.weight  ->  encoder.blocks.16.mlp.0.weight
encoder.blocks.16.mlp.0.weight 2 (5120, 1280)
model.encoder.layers.16.fc1.bias  ->  encoder.blocks.16.mlp.0.bias
encoder.blocks.16.mlp.0.bias 1 (5120,)
  Converting to float32
model.encoder.layers.16.fc2.weight  ->  encoder.blocks.16.mlp.2.weight
encoder.blocks.16.mlp.2.weight 2 (1280, 5120)
model.encoder.layers.16.fc2.bias  ->  encoder.blocks.16.mlp.2.bias
encoder.blocks.16.mlp.2.bias 1 (1280,)
  Converting to float32
model.encoder.layers.16.final_layer_norm.weight  ->  encoder.blocks.16.mlp_ln.weight
encoder.blocks.16.mlp_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.16.final_layer_norm.bias  ->  encoder.blocks.16.mlp_ln.bias
encoder.blocks.16.mlp_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.17.self_attn.k_proj.weight  ->  encoder.blocks.17.attn.key.weight
encoder.blocks.17.attn.key.weight 2 (1280, 1280)
model.encoder.layers.17.self_attn.v_proj.weight  ->  encoder.blocks.17.attn.value.weight
encoder.blocks.17.attn.value.weight 2 (1280, 1280)
model.encoder.layers.17.self_attn.v_proj.bias  ->  encoder.blocks.17.attn.value.bias
encoder.blocks.17.attn.value.bias 1 (1280,)
  Converting to float32
model.encoder.layers.17.self_attn.q_proj.weight  ->  encoder.blocks.17.attn.query.weight
encoder.blocks.17.attn.query.weight 2 (1280, 1280)
model.encoder.layers.17.self_attn.q_proj.bias  ->  encoder.blocks.17.attn.query.bias
encoder.blocks.17.attn.query.bias 1 (1280,)
  Converting to float32
model.encoder.layers.17.self_attn.out_proj.weight  ->  encoder.blocks.17.attn.out.weight
encoder.blocks.17.attn.out.weight 2 (1280, 1280)
model.encoder.layers.17.self_attn.out_proj.bias  ->  encoder.blocks.17.attn.out.bias
encoder.blocks.17.attn.out.bias 1 (1280,)
  Converting to float32
model.encoder.layers.17.self_attn_layer_norm.weight  ->  encoder.blocks.17.attn_ln.weight
encoder.blocks.17.attn_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.17.self_attn_layer_norm.bias  ->  encoder.blocks.17.attn_ln.bias
encoder.blocks.17.attn_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.17.fc1.weight  ->  encoder.blocks.17.mlp.0.weight
encoder.blocks.17.mlp.0.weight 2 (5120, 1280)
model.encoder.layers.17.fc1.bias  ->  encoder.blocks.17.mlp.0.bias
encoder.blocks.17.mlp.0.bias 1 (5120,)
  Converting to float32
model.encoder.layers.17.fc2.weight  ->  encoder.blocks.17.mlp.2.weight
encoder.blocks.17.mlp.2.weight 2 (1280, 5120)
model.encoder.layers.17.fc2.bias  ->  encoder.blocks.17.mlp.2.bias
encoder.blocks.17.mlp.2.bias 1 (1280,)
  Converting to float32
model.encoder.layers.17.final_layer_norm.weight  ->  encoder.blocks.17.mlp_ln.weight
encoder.blocks.17.mlp_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.17.final_layer_norm.bias  ->  encoder.blocks.17.mlp_ln.bias
encoder.blocks.17.mlp_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.18.self_attn.k_proj.weight  ->  encoder.blocks.18.attn.key.weight
encoder.blocks.18.attn.key.weight 2 (1280, 1280)
model.encoder.layers.18.self_attn.v_proj.weight  ->  encoder.blocks.18.attn.value.weight
encoder.blocks.18.attn.value.weight 2 (1280, 1280)
model.encoder.layers.18.self_attn.v_proj.bias  ->  encoder.blocks.18.attn.value.bias
encoder.blocks.18.attn.value.bias 1 (1280,)
  Converting to float32
model.encoder.layers.18.self_attn.q_proj.weight  ->  encoder.blocks.18.attn.query.weight
encoder.blocks.18.attn.query.weight 2 (1280, 1280)
model.encoder.layers.18.self_attn.q_proj.bias  ->  encoder.blocks.18.attn.query.bias
encoder.blocks.18.attn.query.bias 1 (1280,)
  Converting to float32
model.encoder.layers.18.self_attn.out_proj.weight  ->  encoder.blocks.18.attn.out.weight
encoder.blocks.18.attn.out.weight 2 (1280, 1280)
model.encoder.layers.18.self_attn.out_proj.bias  ->  encoder.blocks.18.attn.out.bias
encoder.blocks.18.attn.out.bias 1 (1280,)
  Converting to float32
model.encoder.layers.18.self_attn_layer_norm.weight  ->  encoder.blocks.18.attn_ln.weight
encoder.blocks.18.attn_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.18.self_attn_layer_norm.bias  ->  encoder.blocks.18.attn_ln.bias
encoder.blocks.18.attn_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.18.fc1.weight  ->  encoder.blocks.18.mlp.0.weight
encoder.blocks.18.mlp.0.weight 2 (5120, 1280)
model.encoder.layers.18.fc1.bias  ->  encoder.blocks.18.mlp.0.bias
encoder.blocks.18.mlp.0.bias 1 (5120,)
  Converting to float32
model.encoder.layers.18.fc2.weight  ->  encoder.blocks.18.mlp.2.weight
encoder.blocks.18.mlp.2.weight 2 (1280, 5120)
model.encoder.layers.18.fc2.bias  ->  encoder.blocks.18.mlp.2.bias
encoder.blocks.18.mlp.2.bias 1 (1280,)
  Converting to float32
model.encoder.layers.18.final_layer_norm.weight  ->  encoder.blocks.18.mlp_ln.weight
encoder.blocks.18.mlp_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.18.final_layer_norm.bias  ->  encoder.blocks.18.mlp_ln.bias
encoder.blocks.18.mlp_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.19.self_attn.k_proj.weight  ->  encoder.blocks.19.attn.key.weight
encoder.blocks.19.attn.key.weight 2 (1280, 1280)
model.encoder.layers.19.self_attn.v_proj.weight  ->  encoder.blocks.19.attn.value.weight
encoder.blocks.19.attn.value.weight 2 (1280, 1280)
model.encoder.layers.19.self_attn.v_proj.bias  ->  encoder.blocks.19.attn.value.bias
encoder.blocks.19.attn.value.bias 1 (1280,)
  Converting to float32
model.encoder.layers.19.self_attn.q_proj.weight  ->  encoder.blocks.19.attn.query.weight
encoder.blocks.19.attn.query.weight 2 (1280, 1280)
model.encoder.layers.19.self_attn.q_proj.bias  ->  encoder.blocks.19.attn.query.bias
encoder.blocks.19.attn.query.bias 1 (1280,)
  Converting to float32
model.encoder.layers.19.self_attn.out_proj.weight  ->  encoder.blocks.19.attn.out.weight
encoder.blocks.19.attn.out.weight 2 (1280, 1280)
model.encoder.layers.19.self_attn.out_proj.bias  ->  encoder.blocks.19.attn.out.bias
encoder.blocks.19.attn.out.bias 1 (1280,)
  Converting to float32
model.encoder.layers.19.self_attn_layer_norm.weight  ->  encoder.blocks.19.attn_ln.weight
encoder.blocks.19.attn_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.19.self_attn_layer_norm.bias  ->  encoder.blocks.19.attn_ln.bias
encoder.blocks.19.attn_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.19.fc1.weight  ->  encoder.blocks.19.mlp.0.weight
encoder.blocks.19.mlp.0.weight 2 (5120, 1280)
model.encoder.layers.19.fc1.bias  ->  encoder.blocks.19.mlp.0.bias
encoder.blocks.19.mlp.0.bias 1 (5120,)
  Converting to float32
model.encoder.layers.19.fc2.weight  ->  encoder.blocks.19.mlp.2.weight
encoder.blocks.19.mlp.2.weight 2 (1280, 5120)
model.encoder.layers.19.fc2.bias  ->  encoder.blocks.19.mlp.2.bias
encoder.blocks.19.mlp.2.bias 1 (1280,)
  Converting to float32
model.encoder.layers.19.final_layer_norm.weight  ->  encoder.blocks.19.mlp_ln.weight
encoder.blocks.19.mlp_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.19.final_layer_norm.bias  ->  encoder.blocks.19.mlp_ln.bias
encoder.blocks.19.mlp_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.20.self_attn.k_proj.weight  ->  encoder.blocks.20.attn.key.weight
encoder.blocks.20.attn.key.weight 2 (1280, 1280)
model.encoder.layers.20.self_attn.v_proj.weight  ->  encoder.blocks.20.attn.value.weight
encoder.blocks.20.attn.value.weight 2 (1280, 1280)
model.encoder.layers.20.self_attn.v_proj.bias  ->  encoder.blocks.20.attn.value.bias
encoder.blocks.20.attn.value.bias 1 (1280,)
  Converting to float32
model.encoder.layers.20.self_attn.q_proj.weight  ->  encoder.blocks.20.attn.query.weight
encoder.blocks.20.attn.query.weight 2 (1280, 1280)
model.encoder.layers.20.self_attn.q_proj.bias  ->  encoder.blocks.20.attn.query.bias
encoder.blocks.20.attn.query.bias 1 (1280,)
  Converting to float32
model.encoder.layers.20.self_attn.out_proj.weight  ->  encoder.blocks.20.attn.out.weight
encoder.blocks.20.attn.out.weight 2 (1280, 1280)
model.encoder.layers.20.self_attn.out_proj.bias  ->  encoder.blocks.20.attn.out.bias
encoder.blocks.20.attn.out.bias 1 (1280,)
  Converting to float32
model.encoder.layers.20.self_attn_layer_norm.weight  ->  encoder.blocks.20.attn_ln.weight
encoder.blocks.20.attn_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.20.self_attn_layer_norm.bias  ->  encoder.blocks.20.attn_ln.bias
encoder.blocks.20.attn_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.20.fc1.weight  ->  encoder.blocks.20.mlp.0.weight
encoder.blocks.20.mlp.0.weight 2 (5120, 1280)
model.encoder.layers.20.fc1.bias  ->  encoder.blocks.20.mlp.0.bias
encoder.blocks.20.mlp.0.bias 1 (5120,)
  Converting to float32
model.encoder.layers.20.fc2.weight  ->  encoder.blocks.20.mlp.2.weight
encoder.blocks.20.mlp.2.weight 2 (1280, 5120)
model.encoder.layers.20.fc2.bias  ->  encoder.blocks.20.mlp.2.bias
encoder.blocks.20.mlp.2.bias 1 (1280,)
  Converting to float32
model.encoder.layers.20.final_layer_norm.weight  ->  encoder.blocks.20.mlp_ln.weight
encoder.blocks.20.mlp_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.20.final_layer_norm.bias  ->  encoder.blocks.20.mlp_ln.bias
encoder.blocks.20.mlp_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.21.self_attn.k_proj.weight  ->  encoder.blocks.21.attn.key.weight
encoder.blocks.21.attn.key.weight 2 (1280, 1280)
model.encoder.layers.21.self_attn.v_proj.weight  ->  encoder.blocks.21.attn.value.weight
encoder.blocks.21.attn.value.weight 2 (1280, 1280)
model.encoder.layers.21.self_attn.v_proj.bias  ->  encoder.blocks.21.attn.value.bias
encoder.blocks.21.attn.value.bias 1 (1280,)
  Converting to float32
model.encoder.layers.21.self_attn.q_proj.weight  ->  encoder.blocks.21.attn.query.weight
encoder.blocks.21.attn.query.weight 2 (1280, 1280)
model.encoder.layers.21.self_attn.q_proj.bias  ->  encoder.blocks.21.attn.query.bias
encoder.blocks.21.attn.query.bias 1 (1280,)
  Converting to float32
model.encoder.layers.21.self_attn.out_proj.weight  ->  encoder.blocks.21.attn.out.weight
encoder.blocks.21.attn.out.weight 2 (1280, 1280)
model.encoder.layers.21.self_attn.out_proj.bias  ->  encoder.blocks.21.attn.out.bias
encoder.blocks.21.attn.out.bias 1 (1280,)
  Converting to float32
model.encoder.layers.21.self_attn_layer_norm.weight  ->  encoder.blocks.21.attn_ln.weight
encoder.blocks.21.attn_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.21.self_attn_layer_norm.bias  ->  encoder.blocks.21.attn_ln.bias
encoder.blocks.21.attn_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.21.fc1.weight  ->  encoder.blocks.21.mlp.0.weight
encoder.blocks.21.mlp.0.weight 2 (5120, 1280)
model.encoder.layers.21.fc1.bias  ->  encoder.blocks.21.mlp.0.bias
encoder.blocks.21.mlp.0.bias 1 (5120,)
  Converting to float32
model.encoder.layers.21.fc2.weight  ->  encoder.blocks.21.mlp.2.weight
encoder.blocks.21.mlp.2.weight 2 (1280, 5120)
model.encoder.layers.21.fc2.bias  ->  encoder.blocks.21.mlp.2.bias
encoder.blocks.21.mlp.2.bias 1 (1280,)
  Converting to float32
model.encoder.layers.21.final_layer_norm.weight  ->  encoder.blocks.21.mlp_ln.weight
encoder.blocks.21.mlp_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.21.final_layer_norm.bias  ->  encoder.blocks.21.mlp_ln.bias
encoder.blocks.21.mlp_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.22.self_attn.k_proj.weight  ->  encoder.blocks.22.attn.key.weight
encoder.blocks.22.attn.key.weight 2 (1280, 1280)
model.encoder.layers.22.self_attn.v_proj.weight  ->  encoder.blocks.22.attn.value.weight
encoder.blocks.22.attn.value.weight 2 (1280, 1280)
model.encoder.layers.22.self_attn.v_proj.bias  ->  encoder.blocks.22.attn.value.bias
encoder.blocks.22.attn.value.bias 1 (1280,)
  Converting to float32
model.encoder.layers.22.self_attn.q_proj.weight  ->  encoder.blocks.22.attn.query.weight
encoder.blocks.22.attn.query.weight 2 (1280, 1280)
model.encoder.layers.22.self_attn.q_proj.bias  ->  encoder.blocks.22.attn.query.bias
encoder.blocks.22.attn.query.bias 1 (1280,)
  Converting to float32
model.encoder.layers.22.self_attn.out_proj.weight  ->  encoder.blocks.22.attn.out.weight
encoder.blocks.22.attn.out.weight 2 (1280, 1280)
model.encoder.layers.22.self_attn.out_proj.bias  ->  encoder.blocks.22.attn.out.bias
encoder.blocks.22.attn.out.bias 1 (1280,)
  Converting to float32
model.encoder.layers.22.self_attn_layer_norm.weight  ->  encoder.blocks.22.attn_ln.weight
encoder.blocks.22.attn_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.22.self_attn_layer_norm.bias  ->  encoder.blocks.22.attn_ln.bias
encoder.blocks.22.attn_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.22.fc1.weight  ->  encoder.blocks.22.mlp.0.weight
encoder.blocks.22.mlp.0.weight 2 (5120, 1280)
model.encoder.layers.22.fc1.bias  ->  encoder.blocks.22.mlp.0.bias
encoder.blocks.22.mlp.0.bias 1 (5120,)
  Converting to float32
model.encoder.layers.22.fc2.weight  ->  encoder.blocks.22.mlp.2.weight
encoder.blocks.22.mlp.2.weight 2 (1280, 5120)
model.encoder.layers.22.fc2.bias  ->  encoder.blocks.22.mlp.2.bias
encoder.blocks.22.mlp.2.bias 1 (1280,)
  Converting to float32
model.encoder.layers.22.final_layer_norm.weight  ->  encoder.blocks.22.mlp_ln.weight
encoder.blocks.22.mlp_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.22.final_layer_norm.bias  ->  encoder.blocks.22.mlp_ln.bias
encoder.blocks.22.mlp_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.23.self_attn.k_proj.weight  ->  encoder.blocks.23.attn.key.weight
encoder.blocks.23.attn.key.weight 2 (1280, 1280)
model.encoder.layers.23.self_attn.v_proj.weight  ->  encoder.blocks.23.attn.value.weight
encoder.blocks.23.attn.value.weight 2 (1280, 1280)
model.encoder.layers.23.self_attn.v_proj.bias  ->  encoder.blocks.23.attn.value.bias
encoder.blocks.23.attn.value.bias 1 (1280,)
  Converting to float32
model.encoder.layers.23.self_attn.q_proj.weight  ->  encoder.blocks.23.attn.query.weight
encoder.blocks.23.attn.query.weight 2 (1280, 1280)
model.encoder.layers.23.self_attn.q_proj.bias  ->  encoder.blocks.23.attn.query.bias
encoder.blocks.23.attn.query.bias 1 (1280,)
  Converting to float32
model.encoder.layers.23.self_attn.out_proj.weight  ->  encoder.blocks.23.attn.out.weight
encoder.blocks.23.attn.out.weight 2 (1280, 1280)
model.encoder.layers.23.self_attn.out_proj.bias  ->  encoder.blocks.23.attn.out.bias
encoder.blocks.23.attn.out.bias 1 (1280,)
  Converting to float32
model.encoder.layers.23.self_attn_layer_norm.weight  ->  encoder.blocks.23.attn_ln.weight
encoder.blocks.23.attn_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.23.self_attn_layer_norm.bias  ->  encoder.blocks.23.attn_ln.bias
encoder.blocks.23.attn_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.23.fc1.weight  ->  encoder.blocks.23.mlp.0.weight
encoder.blocks.23.mlp.0.weight 2 (5120, 1280)
model.encoder.layers.23.fc1.bias  ->  encoder.blocks.23.mlp.0.bias
encoder.blocks.23.mlp.0.bias 1 (5120,)
  Converting to float32
model.encoder.layers.23.fc2.weight  ->  encoder.blocks.23.mlp.2.weight
encoder.blocks.23.mlp.2.weight 2 (1280, 5120)
model.encoder.layers.23.fc2.bias  ->  encoder.blocks.23.mlp.2.bias
encoder.blocks.23.mlp.2.bias 1 (1280,)
  Converting to float32
model.encoder.layers.23.final_layer_norm.weight  ->  encoder.blocks.23.mlp_ln.weight
encoder.blocks.23.mlp_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.23.final_layer_norm.bias  ->  encoder.blocks.23.mlp_ln.bias
encoder.blocks.23.mlp_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.24.self_attn.k_proj.weight  ->  encoder.blocks.24.attn.key.weight
encoder.blocks.24.attn.key.weight 2 (1280, 1280)
model.encoder.layers.24.self_attn.v_proj.weight  ->  encoder.blocks.24.attn.value.weight
encoder.blocks.24.attn.value.weight 2 (1280, 1280)
model.encoder.layers.24.self_attn.v_proj.bias  ->  encoder.blocks.24.attn.value.bias
encoder.blocks.24.attn.value.bias 1 (1280,)
  Converting to float32
model.encoder.layers.24.self_attn.q_proj.weight  ->  encoder.blocks.24.attn.query.weight
encoder.blocks.24.attn.query.weight 2 (1280, 1280)
model.encoder.layers.24.self_attn.q_proj.bias  ->  encoder.blocks.24.attn.query.bias
encoder.blocks.24.attn.query.bias 1 (1280,)
  Converting to float32
model.encoder.layers.24.self_attn.out_proj.weight  ->  encoder.blocks.24.attn.out.weight
encoder.blocks.24.attn.out.weight 2 (1280, 1280)
model.encoder.layers.24.self_attn.out_proj.bias  ->  encoder.blocks.24.attn.out.bias
encoder.blocks.24.attn.out.bias 1 (1280,)
  Converting to float32
model.encoder.layers.24.self_attn_layer_norm.weight  ->  encoder.blocks.24.attn_ln.weight
encoder.blocks.24.attn_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.24.self_attn_layer_norm.bias  ->  encoder.blocks.24.attn_ln.bias
encoder.blocks.24.attn_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.24.fc1.weight  ->  encoder.blocks.24.mlp.0.weight
encoder.blocks.24.mlp.0.weight 2 (5120, 1280)
model.encoder.layers.24.fc1.bias  ->  encoder.blocks.24.mlp.0.bias
encoder.blocks.24.mlp.0.bias 1 (5120,)
  Converting to float32
model.encoder.layers.24.fc2.weight  ->  encoder.blocks.24.mlp.2.weight
encoder.blocks.24.mlp.2.weight 2 (1280, 5120)
model.encoder.layers.24.fc2.bias  ->  encoder.blocks.24.mlp.2.bias
encoder.blocks.24.mlp.2.bias 1 (1280,)
  Converting to float32
model.encoder.layers.24.final_layer_norm.weight  ->  encoder.blocks.24.mlp_ln.weight
encoder.blocks.24.mlp_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.24.final_layer_norm.bias  ->  encoder.blocks.24.mlp_ln.bias
encoder.blocks.24.mlp_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.25.self_attn.k_proj.weight  ->  encoder.blocks.25.attn.key.weight
encoder.blocks.25.attn.key.weight 2 (1280, 1280)
model.encoder.layers.25.self_attn.v_proj.weight  ->  encoder.blocks.25.attn.value.weight
encoder.blocks.25.attn.value.weight 2 (1280, 1280)
model.encoder.layers.25.self_attn.v_proj.bias  ->  encoder.blocks.25.attn.value.bias
encoder.blocks.25.attn.value.bias 1 (1280,)
  Converting to float32
model.encoder.layers.25.self_attn.q_proj.weight  ->  encoder.blocks.25.attn.query.weight
encoder.blocks.25.attn.query.weight 2 (1280, 1280)
model.encoder.layers.25.self_attn.q_proj.bias  ->  encoder.blocks.25.attn.query.bias
encoder.blocks.25.attn.query.bias 1 (1280,)
  Converting to float32
model.encoder.layers.25.self_attn.out_proj.weight  ->  encoder.blocks.25.attn.out.weight
encoder.blocks.25.attn.out.weight 2 (1280, 1280)
model.encoder.layers.25.self_attn.out_proj.bias  ->  encoder.blocks.25.attn.out.bias
encoder.blocks.25.attn.out.bias 1 (1280,)
  Converting to float32
model.encoder.layers.25.self_attn_layer_norm.weight  ->  encoder.blocks.25.attn_ln.weight
encoder.blocks.25.attn_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.25.self_attn_layer_norm.bias  ->  encoder.blocks.25.attn_ln.bias
encoder.blocks.25.attn_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.25.fc1.weight  ->  encoder.blocks.25.mlp.0.weight
encoder.blocks.25.mlp.0.weight 2 (5120, 1280)
model.encoder.layers.25.fc1.bias  ->  encoder.blocks.25.mlp.0.bias
encoder.blocks.25.mlp.0.bias 1 (5120,)
  Converting to float32
model.encoder.layers.25.fc2.weight  ->  encoder.blocks.25.mlp.2.weight
encoder.blocks.25.mlp.2.weight 2 (1280, 5120)
model.encoder.layers.25.fc2.bias  ->  encoder.blocks.25.mlp.2.bias
encoder.blocks.25.mlp.2.bias 1 (1280,)
  Converting to float32
model.encoder.layers.25.final_layer_norm.weight  ->  encoder.blocks.25.mlp_ln.weight
encoder.blocks.25.mlp_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.25.final_layer_norm.bias  ->  encoder.blocks.25.mlp_ln.bias
encoder.blocks.25.mlp_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.26.self_attn.k_proj.weight  ->  encoder.blocks.26.attn.key.weight
encoder.blocks.26.attn.key.weight 2 (1280, 1280)
model.encoder.layers.26.self_attn.v_proj.weight  ->  encoder.blocks.26.attn.value.weight
encoder.blocks.26.attn.value.weight 2 (1280, 1280)
model.encoder.layers.26.self_attn.v_proj.bias  ->  encoder.blocks.26.attn.value.bias
encoder.blocks.26.attn.value.bias 1 (1280,)
  Converting to float32
model.encoder.layers.26.self_attn.q_proj.weight  ->  encoder.blocks.26.attn.query.weight
encoder.blocks.26.attn.query.weight 2 (1280, 1280)
model.encoder.layers.26.self_attn.q_proj.bias  ->  encoder.blocks.26.attn.query.bias
encoder.blocks.26.attn.query.bias 1 (1280,)
  Converting to float32
model.encoder.layers.26.self_attn.out_proj.weight  ->  encoder.blocks.26.attn.out.weight
encoder.blocks.26.attn.out.weight 2 (1280, 1280)
model.encoder.layers.26.self_attn.out_proj.bias  ->  encoder.blocks.26.attn.out.bias
encoder.blocks.26.attn.out.bias 1 (1280,)
  Converting to float32
model.encoder.layers.26.self_attn_layer_norm.weight  ->  encoder.blocks.26.attn_ln.weight
encoder.blocks.26.attn_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.26.self_attn_layer_norm.bias  ->  encoder.blocks.26.attn_ln.bias
encoder.blocks.26.attn_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.26.fc1.weight  ->  encoder.blocks.26.mlp.0.weight
encoder.blocks.26.mlp.0.weight 2 (5120, 1280)
model.encoder.layers.26.fc1.bias  ->  encoder.blocks.26.mlp.0.bias
encoder.blocks.26.mlp.0.bias 1 (5120,)
  Converting to float32
model.encoder.layers.26.fc2.weight  ->  encoder.blocks.26.mlp.2.weight
encoder.blocks.26.mlp.2.weight 2 (1280, 5120)
model.encoder.layers.26.fc2.bias  ->  encoder.blocks.26.mlp.2.bias
encoder.blocks.26.mlp.2.bias 1 (1280,)
  Converting to float32
model.encoder.layers.26.final_layer_norm.weight  ->  encoder.blocks.26.mlp_ln.weight
encoder.blocks.26.mlp_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.26.final_layer_norm.bias  ->  encoder.blocks.26.mlp_ln.bias
encoder.blocks.26.mlp_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.27.self_attn.k_proj.weight  ->  encoder.blocks.27.attn.key.weight
encoder.blocks.27.attn.key.weight 2 (1280, 1280)
model.encoder.layers.27.self_attn.v_proj.weight  ->  encoder.blocks.27.attn.value.weight
encoder.blocks.27.attn.value.weight 2 (1280, 1280)
model.encoder.layers.27.self_attn.v_proj.bias  ->  encoder.blocks.27.attn.value.bias
encoder.blocks.27.attn.value.bias 1 (1280,)
  Converting to float32
model.encoder.layers.27.self_attn.q_proj.weight  ->  encoder.blocks.27.attn.query.weight
encoder.blocks.27.attn.query.weight 2 (1280, 1280)
model.encoder.layers.27.self_attn.q_proj.bias  ->  encoder.blocks.27.attn.query.bias
encoder.blocks.27.attn.query.bias 1 (1280,)
  Converting to float32
model.encoder.layers.27.self_attn.out_proj.weight  ->  encoder.blocks.27.attn.out.weight
encoder.blocks.27.attn.out.weight 2 (1280, 1280)
model.encoder.layers.27.self_attn.out_proj.bias  ->  encoder.blocks.27.attn.out.bias
encoder.blocks.27.attn.out.bias 1 (1280,)
  Converting to float32
model.encoder.layers.27.self_attn_layer_norm.weight  ->  encoder.blocks.27.attn_ln.weight
encoder.blocks.27.attn_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.27.self_attn_layer_norm.bias  ->  encoder.blocks.27.attn_ln.bias
encoder.blocks.27.attn_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.27.fc1.weight  ->  encoder.blocks.27.mlp.0.weight
encoder.blocks.27.mlp.0.weight 2 (5120, 1280)
model.encoder.layers.27.fc1.bias  ->  encoder.blocks.27.mlp.0.bias
encoder.blocks.27.mlp.0.bias 1 (5120,)
  Converting to float32
model.encoder.layers.27.fc2.weight  ->  encoder.blocks.27.mlp.2.weight
encoder.blocks.27.mlp.2.weight 2 (1280, 5120)
model.encoder.layers.27.fc2.bias  ->  encoder.blocks.27.mlp.2.bias
encoder.blocks.27.mlp.2.bias 1 (1280,)
  Converting to float32
model.encoder.layers.27.final_layer_norm.weight  ->  encoder.blocks.27.mlp_ln.weight
encoder.blocks.27.mlp_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.27.final_layer_norm.bias  ->  encoder.blocks.27.mlp_ln.bias
encoder.blocks.27.mlp_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.28.self_attn.k_proj.weight  ->  encoder.blocks.28.attn.key.weight
encoder.blocks.28.attn.key.weight 2 (1280, 1280)
model.encoder.layers.28.self_attn.v_proj.weight  ->  encoder.blocks.28.attn.value.weight
encoder.blocks.28.attn.value.weight 2 (1280, 1280)
model.encoder.layers.28.self_attn.v_proj.bias  ->  encoder.blocks.28.attn.value.bias
encoder.blocks.28.attn.value.bias 1 (1280,)
  Converting to float32
model.encoder.layers.28.self_attn.q_proj.weight  ->  encoder.blocks.28.attn.query.weight
encoder.blocks.28.attn.query.weight 2 (1280, 1280)
model.encoder.layers.28.self_attn.q_proj.bias  ->  encoder.blocks.28.attn.query.bias
encoder.blocks.28.attn.query.bias 1 (1280,)
  Converting to float32
model.encoder.layers.28.self_attn.out_proj.weight  ->  encoder.blocks.28.attn.out.weight
encoder.blocks.28.attn.out.weight 2 (1280, 1280)
model.encoder.layers.28.self_attn.out_proj.bias  ->  encoder.blocks.28.attn.out.bias
encoder.blocks.28.attn.out.bias 1 (1280,)
  Converting to float32
model.encoder.layers.28.self_attn_layer_norm.weight  ->  encoder.blocks.28.attn_ln.weight
encoder.blocks.28.attn_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.28.self_attn_layer_norm.bias  ->  encoder.blocks.28.attn_ln.bias
encoder.blocks.28.attn_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.28.fc1.weight  ->  encoder.blocks.28.mlp.0.weight
encoder.blocks.28.mlp.0.weight 2 (5120, 1280)
model.encoder.layers.28.fc1.bias  ->  encoder.blocks.28.mlp.0.bias
encoder.blocks.28.mlp.0.bias 1 (5120,)
  Converting to float32
model.encoder.layers.28.fc2.weight  ->  encoder.blocks.28.mlp.2.weight
encoder.blocks.28.mlp.2.weight 2 (1280, 5120)
model.encoder.layers.28.fc2.bias  ->  encoder.blocks.28.mlp.2.bias
encoder.blocks.28.mlp.2.bias 1 (1280,)
  Converting to float32
model.encoder.layers.28.final_layer_norm.weight  ->  encoder.blocks.28.mlp_ln.weight
encoder.blocks.28.mlp_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.28.final_layer_norm.bias  ->  encoder.blocks.28.mlp_ln.bias
encoder.blocks.28.mlp_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.29.self_attn.k_proj.weight  ->  encoder.blocks.29.attn.key.weight
encoder.blocks.29.attn.key.weight 2 (1280, 1280)
model.encoder.layers.29.self_attn.v_proj.weight  ->  encoder.blocks.29.attn.value.weight
encoder.blocks.29.attn.value.weight 2 (1280, 1280)
model.encoder.layers.29.self_attn.v_proj.bias  ->  encoder.blocks.29.attn.value.bias
encoder.blocks.29.attn.value.bias 1 (1280,)
  Converting to float32
model.encoder.layers.29.self_attn.q_proj.weight  ->  encoder.blocks.29.attn.query.weight
encoder.blocks.29.attn.query.weight 2 (1280, 1280)
model.encoder.layers.29.self_attn.q_proj.bias  ->  encoder.blocks.29.attn.query.bias
encoder.blocks.29.attn.query.bias 1 (1280,)
  Converting to float32
model.encoder.layers.29.self_attn.out_proj.weight  ->  encoder.blocks.29.attn.out.weight
encoder.blocks.29.attn.out.weight 2 (1280, 1280)
model.encoder.layers.29.self_attn.out_proj.bias  ->  encoder.blocks.29.attn.out.bias
encoder.blocks.29.attn.out.bias 1 (1280,)
  Converting to float32
model.encoder.layers.29.self_attn_layer_norm.weight  ->  encoder.blocks.29.attn_ln.weight
encoder.blocks.29.attn_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.29.self_attn_layer_norm.bias  ->  encoder.blocks.29.attn_ln.bias
encoder.blocks.29.attn_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.29.fc1.weight  ->  encoder.blocks.29.mlp.0.weight
encoder.blocks.29.mlp.0.weight 2 (5120, 1280)
model.encoder.layers.29.fc1.bias  ->  encoder.blocks.29.mlp.0.bias
encoder.blocks.29.mlp.0.bias 1 (5120,)
  Converting to float32
model.encoder.layers.29.fc2.weight  ->  encoder.blocks.29.mlp.2.weight
encoder.blocks.29.mlp.2.weight 2 (1280, 5120)
model.encoder.layers.29.fc2.bias  ->  encoder.blocks.29.mlp.2.bias
encoder.blocks.29.mlp.2.bias 1 (1280,)
  Converting to float32
model.encoder.layers.29.final_layer_norm.weight  ->  encoder.blocks.29.mlp_ln.weight
encoder.blocks.29.mlp_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.29.final_layer_norm.bias  ->  encoder.blocks.29.mlp_ln.bias
encoder.blocks.29.mlp_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.30.self_attn.k_proj.weight  ->  encoder.blocks.30.attn.key.weight
encoder.blocks.30.attn.key.weight 2 (1280, 1280)
model.encoder.layers.30.self_attn.v_proj.weight  ->  encoder.blocks.30.attn.value.weight
encoder.blocks.30.attn.value.weight 2 (1280, 1280)
model.encoder.layers.30.self_attn.v_proj.bias  ->  encoder.blocks.30.attn.value.bias
encoder.blocks.30.attn.value.bias 1 (1280,)
  Converting to float32
model.encoder.layers.30.self_attn.q_proj.weight  ->  encoder.blocks.30.attn.query.weight
encoder.blocks.30.attn.query.weight 2 (1280, 1280)
model.encoder.layers.30.self_attn.q_proj.bias  ->  encoder.blocks.30.attn.query.bias
encoder.blocks.30.attn.query.bias 1 (1280,)
  Converting to float32
model.encoder.layers.30.self_attn.out_proj.weight  ->  encoder.blocks.30.attn.out.weight
encoder.blocks.30.attn.out.weight 2 (1280, 1280)
model.encoder.layers.30.self_attn.out_proj.bias  ->  encoder.blocks.30.attn.out.bias
encoder.blocks.30.attn.out.bias 1 (1280,)
  Converting to float32
model.encoder.layers.30.self_attn_layer_norm.weight  ->  encoder.blocks.30.attn_ln.weight
encoder.blocks.30.attn_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.30.self_attn_layer_norm.bias  ->  encoder.blocks.30.attn_ln.bias
encoder.blocks.30.attn_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.30.fc1.weight  ->  encoder.blocks.30.mlp.0.weight
encoder.blocks.30.mlp.0.weight 2 (5120, 1280)
model.encoder.layers.30.fc1.bias  ->  encoder.blocks.30.mlp.0.bias
encoder.blocks.30.mlp.0.bias 1 (5120,)
  Converting to float32
model.encoder.layers.30.fc2.weight  ->  encoder.blocks.30.mlp.2.weight
encoder.blocks.30.mlp.2.weight 2 (1280, 5120)
model.encoder.layers.30.fc2.bias  ->  encoder.blocks.30.mlp.2.bias
encoder.blocks.30.mlp.2.bias 1 (1280,)
  Converting to float32
model.encoder.layers.30.final_layer_norm.weight  ->  encoder.blocks.30.mlp_ln.weight
encoder.blocks.30.mlp_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.30.final_layer_norm.bias  ->  encoder.blocks.30.mlp_ln.bias
encoder.blocks.30.mlp_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.31.self_attn.k_proj.weight  ->  encoder.blocks.31.attn.key.weight
encoder.blocks.31.attn.key.weight 2 (1280, 1280)
model.encoder.layers.31.self_attn.v_proj.weight  ->  encoder.blocks.31.attn.value.weight
encoder.blocks.31.attn.value.weight 2 (1280, 1280)
model.encoder.layers.31.self_attn.v_proj.bias  ->  encoder.blocks.31.attn.value.bias
encoder.blocks.31.attn.value.bias 1 (1280,)
  Converting to float32
model.encoder.layers.31.self_attn.q_proj.weight  ->  encoder.blocks.31.attn.query.weight
encoder.blocks.31.attn.query.weight 2 (1280, 1280)
model.encoder.layers.31.self_attn.q_proj.bias  ->  encoder.blocks.31.attn.query.bias
encoder.blocks.31.attn.query.bias 1 (1280,)
  Converting to float32
model.encoder.layers.31.self_attn.out_proj.weight  ->  encoder.blocks.31.attn.out.weight
encoder.blocks.31.attn.out.weight 2 (1280, 1280)
model.encoder.layers.31.self_attn.out_proj.bias  ->  encoder.blocks.31.attn.out.bias
encoder.blocks.31.attn.out.bias 1 (1280,)
  Converting to float32
model.encoder.layers.31.self_attn_layer_norm.weight  ->  encoder.blocks.31.attn_ln.weight
encoder.blocks.31.attn_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.31.self_attn_layer_norm.bias  ->  encoder.blocks.31.attn_ln.bias
encoder.blocks.31.attn_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layers.31.fc1.weight  ->  encoder.blocks.31.mlp.0.weight
encoder.blocks.31.mlp.0.weight 2 (5120, 1280)
model.encoder.layers.31.fc1.bias  ->  encoder.blocks.31.mlp.0.bias
encoder.blocks.31.mlp.0.bias 1 (5120,)
  Converting to float32
model.encoder.layers.31.fc2.weight  ->  encoder.blocks.31.mlp.2.weight
encoder.blocks.31.mlp.2.weight 2 (1280, 5120)
model.encoder.layers.31.fc2.bias  ->  encoder.blocks.31.mlp.2.bias
encoder.blocks.31.mlp.2.bias 1 (1280,)
  Converting to float32
model.encoder.layers.31.final_layer_norm.weight  ->  encoder.blocks.31.mlp_ln.weight
encoder.blocks.31.mlp_ln.weight 1 (1280,)
  Converting to float32
model.encoder.layers.31.final_layer_norm.bias  ->  encoder.blocks.31.mlp_ln.bias
encoder.blocks.31.mlp_ln.bias 1 (1280,)
  Converting to float32
model.encoder.layer_norm.weight  ->  encoder.ln_post.weight
encoder.ln_post.weight 1 (1280,)
  Converting to float32
model.encoder.layer_norm.bias  ->  encoder.ln_post.bias
encoder.ln_post.bias 1 (1280,)
  Converting to float32
model.decoder.embed_tokens.weight  ->  decoder.token_embedding.weight
decoder.token_embedding.weight 2 (51865, 1280)
model.decoder.embed_positions.weight  ->  decoder.positional_embedding
decoder.positional_embedding 2 (448, 1280)
  Converting to float32
model.decoder.layers.0.self_attn.k_proj.weight  ->  decoder.blocks.0.attn.key.weight
decoder.blocks.0.attn.key.weight 2 (1280, 1280)
model.decoder.layers.0.self_attn.v_proj.weight  ->  decoder.blocks.0.attn.value.weight
decoder.blocks.0.attn.value.weight 2 (1280, 1280)
model.decoder.layers.0.self_attn.v_proj.bias  ->  decoder.blocks.0.attn.value.bias
decoder.blocks.0.attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.0.self_attn.q_proj.weight  ->  decoder.blocks.0.attn.query.weight
decoder.blocks.0.attn.query.weight 2 (1280, 1280)
model.decoder.layers.0.self_attn.q_proj.bias  ->  decoder.blocks.0.attn.query.bias
decoder.blocks.0.attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.0.self_attn.out_proj.weight  ->  decoder.blocks.0.attn.out.weight
decoder.blocks.0.attn.out.weight 2 (1280, 1280)
model.decoder.layers.0.self_attn.out_proj.bias  ->  decoder.blocks.0.attn.out.bias
decoder.blocks.0.attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.0.self_attn_layer_norm.weight  ->  decoder.blocks.0.attn_ln.weight
decoder.blocks.0.attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.0.self_attn_layer_norm.bias  ->  decoder.blocks.0.attn_ln.bias
decoder.blocks.0.attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.0.encoder_attn.k_proj.weight  ->  decoder.blocks.0.cross_attn.key.weight
decoder.blocks.0.cross_attn.key.weight 2 (1280, 1280)
model.decoder.layers.0.encoder_attn.v_proj.weight  ->  decoder.blocks.0.cross_attn.value.weight
decoder.blocks.0.cross_attn.value.weight 2 (1280, 1280)
model.decoder.layers.0.encoder_attn.v_proj.bias  ->  decoder.blocks.0.cross_attn.value.bias
decoder.blocks.0.cross_attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.0.encoder_attn.q_proj.weight  ->  decoder.blocks.0.cross_attn.query.weight
decoder.blocks.0.cross_attn.query.weight 2 (1280, 1280)
model.decoder.layers.0.encoder_attn.q_proj.bias  ->  decoder.blocks.0.cross_attn.query.bias
decoder.blocks.0.cross_attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.0.encoder_attn.out_proj.weight  ->  decoder.blocks.0.cross_attn.out.weight
decoder.blocks.0.cross_attn.out.weight 2 (1280, 1280)
model.decoder.layers.0.encoder_attn.out_proj.bias  ->  decoder.blocks.0.cross_attn.out.bias
decoder.blocks.0.cross_attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.0.encoder_attn_layer_norm.weight  ->  decoder.blocks.0.cross_attn_ln.weight
decoder.blocks.0.cross_attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.0.encoder_attn_layer_norm.bias  ->  decoder.blocks.0.cross_attn_ln.bias
decoder.blocks.0.cross_attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.0.fc1.weight  ->  decoder.blocks.0.mlp.0.weight
decoder.blocks.0.mlp.0.weight 2 (5120, 1280)
model.decoder.layers.0.fc1.bias  ->  decoder.blocks.0.mlp.0.bias
decoder.blocks.0.mlp.0.bias 1 (5120,)
  Converting to float32
model.decoder.layers.0.fc2.weight  ->  decoder.blocks.0.mlp.2.weight
decoder.blocks.0.mlp.2.weight 2 (1280, 5120)
model.decoder.layers.0.fc2.bias  ->  decoder.blocks.0.mlp.2.bias
decoder.blocks.0.mlp.2.bias 1 (1280,)
  Converting to float32
model.decoder.layers.0.final_layer_norm.weight  ->  decoder.blocks.0.mlp_ln.weight
decoder.blocks.0.mlp_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.0.final_layer_norm.bias  ->  decoder.blocks.0.mlp_ln.bias
decoder.blocks.0.mlp_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.1.self_attn.k_proj.weight  ->  decoder.blocks.1.attn.key.weight
decoder.blocks.1.attn.key.weight 2 (1280, 1280)
model.decoder.layers.1.self_attn.v_proj.weight  ->  decoder.blocks.1.attn.value.weight
decoder.blocks.1.attn.value.weight 2 (1280, 1280)
model.decoder.layers.1.self_attn.v_proj.bias  ->  decoder.blocks.1.attn.value.bias
decoder.blocks.1.attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.1.self_attn.q_proj.weight  ->  decoder.blocks.1.attn.query.weight
decoder.blocks.1.attn.query.weight 2 (1280, 1280)
model.decoder.layers.1.self_attn.q_proj.bias  ->  decoder.blocks.1.attn.query.bias
decoder.blocks.1.attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.1.self_attn.out_proj.weight  ->  decoder.blocks.1.attn.out.weight
decoder.blocks.1.attn.out.weight 2 (1280, 1280)
model.decoder.layers.1.self_attn.out_proj.bias  ->  decoder.blocks.1.attn.out.bias
decoder.blocks.1.attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.1.self_attn_layer_norm.weight  ->  decoder.blocks.1.attn_ln.weight
decoder.blocks.1.attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.1.self_attn_layer_norm.bias  ->  decoder.blocks.1.attn_ln.bias
decoder.blocks.1.attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.1.encoder_attn.k_proj.weight  ->  decoder.blocks.1.cross_attn.key.weight
decoder.blocks.1.cross_attn.key.weight 2 (1280, 1280)
model.decoder.layers.1.encoder_attn.v_proj.weight  ->  decoder.blocks.1.cross_attn.value.weight
decoder.blocks.1.cross_attn.value.weight 2 (1280, 1280)
model.decoder.layers.1.encoder_attn.v_proj.bias  ->  decoder.blocks.1.cross_attn.value.bias
decoder.blocks.1.cross_attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.1.encoder_attn.q_proj.weight  ->  decoder.blocks.1.cross_attn.query.weight
decoder.blocks.1.cross_attn.query.weight 2 (1280, 1280)
model.decoder.layers.1.encoder_attn.q_proj.bias  ->  decoder.blocks.1.cross_attn.query.bias
decoder.blocks.1.cross_attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.1.encoder_attn.out_proj.weight  ->  decoder.blocks.1.cross_attn.out.weight
decoder.blocks.1.cross_attn.out.weight 2 (1280, 1280)
model.decoder.layers.1.encoder_attn.out_proj.bias  ->  decoder.blocks.1.cross_attn.out.bias
decoder.blocks.1.cross_attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.1.encoder_attn_layer_norm.weight  ->  decoder.blocks.1.cross_attn_ln.weight
decoder.blocks.1.cross_attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.1.encoder_attn_layer_norm.bias  ->  decoder.blocks.1.cross_attn_ln.bias
decoder.blocks.1.cross_attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.1.fc1.weight  ->  decoder.blocks.1.mlp.0.weight
decoder.blocks.1.mlp.0.weight 2 (5120, 1280)
model.decoder.layers.1.fc1.bias  ->  decoder.blocks.1.mlp.0.bias
decoder.blocks.1.mlp.0.bias 1 (5120,)
  Converting to float32
model.decoder.layers.1.fc2.weight  ->  decoder.blocks.1.mlp.2.weight
decoder.blocks.1.mlp.2.weight 2 (1280, 5120)
model.decoder.layers.1.fc2.bias  ->  decoder.blocks.1.mlp.2.bias
decoder.blocks.1.mlp.2.bias 1 (1280,)
  Converting to float32
model.decoder.layers.1.final_layer_norm.weight  ->  decoder.blocks.1.mlp_ln.weight
decoder.blocks.1.mlp_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.1.final_layer_norm.bias  ->  decoder.blocks.1.mlp_ln.bias
decoder.blocks.1.mlp_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.2.self_attn.k_proj.weight  ->  decoder.blocks.2.attn.key.weight
decoder.blocks.2.attn.key.weight 2 (1280, 1280)
model.decoder.layers.2.self_attn.v_proj.weight  ->  decoder.blocks.2.attn.value.weight
decoder.blocks.2.attn.value.weight 2 (1280, 1280)
model.decoder.layers.2.self_attn.v_proj.bias  ->  decoder.blocks.2.attn.value.bias
decoder.blocks.2.attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.2.self_attn.q_proj.weight  ->  decoder.blocks.2.attn.query.weight
decoder.blocks.2.attn.query.weight 2 (1280, 1280)
model.decoder.layers.2.self_attn.q_proj.bias  ->  decoder.blocks.2.attn.query.bias
decoder.blocks.2.attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.2.self_attn.out_proj.weight  ->  decoder.blocks.2.attn.out.weight
decoder.blocks.2.attn.out.weight 2 (1280, 1280)
model.decoder.layers.2.self_attn.out_proj.bias  ->  decoder.blocks.2.attn.out.bias
decoder.blocks.2.attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.2.self_attn_layer_norm.weight  ->  decoder.blocks.2.attn_ln.weight
decoder.blocks.2.attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.2.self_attn_layer_norm.bias  ->  decoder.blocks.2.attn_ln.bias
decoder.blocks.2.attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.2.encoder_attn.k_proj.weight  ->  decoder.blocks.2.cross_attn.key.weight
decoder.blocks.2.cross_attn.key.weight 2 (1280, 1280)
model.decoder.layers.2.encoder_attn.v_proj.weight  ->  decoder.blocks.2.cross_attn.value.weight
decoder.blocks.2.cross_attn.value.weight 2 (1280, 1280)
model.decoder.layers.2.encoder_attn.v_proj.bias  ->  decoder.blocks.2.cross_attn.value.bias
decoder.blocks.2.cross_attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.2.encoder_attn.q_proj.weight  ->  decoder.blocks.2.cross_attn.query.weight
decoder.blocks.2.cross_attn.query.weight 2 (1280, 1280)
model.decoder.layers.2.encoder_attn.q_proj.bias  ->  decoder.blocks.2.cross_attn.query.bias
decoder.blocks.2.cross_attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.2.encoder_attn.out_proj.weight  ->  decoder.blocks.2.cross_attn.out.weight
decoder.blocks.2.cross_attn.out.weight 2 (1280, 1280)
model.decoder.layers.2.encoder_attn.out_proj.bias  ->  decoder.blocks.2.cross_attn.out.bias
decoder.blocks.2.cross_attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.2.encoder_attn_layer_norm.weight  ->  decoder.blocks.2.cross_attn_ln.weight
decoder.blocks.2.cross_attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.2.encoder_attn_layer_norm.bias  ->  decoder.blocks.2.cross_attn_ln.bias
decoder.blocks.2.cross_attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.2.fc1.weight  ->  decoder.blocks.2.mlp.0.weight
decoder.blocks.2.mlp.0.weight 2 (5120, 1280)
model.decoder.layers.2.fc1.bias  ->  decoder.blocks.2.mlp.0.bias
decoder.blocks.2.mlp.0.bias 1 (5120,)
  Converting to float32
model.decoder.layers.2.fc2.weight  ->  decoder.blocks.2.mlp.2.weight
decoder.blocks.2.mlp.2.weight 2 (1280, 5120)
model.decoder.layers.2.fc2.bias  ->  decoder.blocks.2.mlp.2.bias
decoder.blocks.2.mlp.2.bias 1 (1280,)
  Converting to float32
model.decoder.layers.2.final_layer_norm.weight  ->  decoder.blocks.2.mlp_ln.weight
decoder.blocks.2.mlp_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.2.final_layer_norm.bias  ->  decoder.blocks.2.mlp_ln.bias
decoder.blocks.2.mlp_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.3.self_attn.k_proj.weight  ->  decoder.blocks.3.attn.key.weight
decoder.blocks.3.attn.key.weight 2 (1280, 1280)
model.decoder.layers.3.self_attn.v_proj.weight  ->  decoder.blocks.3.attn.value.weight
decoder.blocks.3.attn.value.weight 2 (1280, 1280)
model.decoder.layers.3.self_attn.v_proj.bias  ->  decoder.blocks.3.attn.value.bias
decoder.blocks.3.attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.3.self_attn.q_proj.weight  ->  decoder.blocks.3.attn.query.weight
decoder.blocks.3.attn.query.weight 2 (1280, 1280)
model.decoder.layers.3.self_attn.q_proj.bias  ->  decoder.blocks.3.attn.query.bias
decoder.blocks.3.attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.3.self_attn.out_proj.weight  ->  decoder.blocks.3.attn.out.weight
decoder.blocks.3.attn.out.weight 2 (1280, 1280)
model.decoder.layers.3.self_attn.out_proj.bias  ->  decoder.blocks.3.attn.out.bias
decoder.blocks.3.attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.3.self_attn_layer_norm.weight  ->  decoder.blocks.3.attn_ln.weight
decoder.blocks.3.attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.3.self_attn_layer_norm.bias  ->  decoder.blocks.3.attn_ln.bias
decoder.blocks.3.attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.3.encoder_attn.k_proj.weight  ->  decoder.blocks.3.cross_attn.key.weight
decoder.blocks.3.cross_attn.key.weight 2 (1280, 1280)
model.decoder.layers.3.encoder_attn.v_proj.weight  ->  decoder.blocks.3.cross_attn.value.weight
decoder.blocks.3.cross_attn.value.weight 2 (1280, 1280)
model.decoder.layers.3.encoder_attn.v_proj.bias  ->  decoder.blocks.3.cross_attn.value.bias
decoder.blocks.3.cross_attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.3.encoder_attn.q_proj.weight  ->  decoder.blocks.3.cross_attn.query.weight
decoder.blocks.3.cross_attn.query.weight 2 (1280, 1280)
model.decoder.layers.3.encoder_attn.q_proj.bias  ->  decoder.blocks.3.cross_attn.query.bias
decoder.blocks.3.cross_attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.3.encoder_attn.out_proj.weight  ->  decoder.blocks.3.cross_attn.out.weight
decoder.blocks.3.cross_attn.out.weight 2 (1280, 1280)
model.decoder.layers.3.encoder_attn.out_proj.bias  ->  decoder.blocks.3.cross_attn.out.bias
decoder.blocks.3.cross_attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.3.encoder_attn_layer_norm.weight  ->  decoder.blocks.3.cross_attn_ln.weight
decoder.blocks.3.cross_attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.3.encoder_attn_layer_norm.bias  ->  decoder.blocks.3.cross_attn_ln.bias
decoder.blocks.3.cross_attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.3.fc1.weight  ->  decoder.blocks.3.mlp.0.weight
decoder.blocks.3.mlp.0.weight 2 (5120, 1280)
model.decoder.layers.3.fc1.bias  ->  decoder.blocks.3.mlp.0.bias
decoder.blocks.3.mlp.0.bias 1 (5120,)
  Converting to float32
model.decoder.layers.3.fc2.weight  ->  decoder.blocks.3.mlp.2.weight
decoder.blocks.3.mlp.2.weight 2 (1280, 5120)
model.decoder.layers.3.fc2.bias  ->  decoder.blocks.3.mlp.2.bias
decoder.blocks.3.mlp.2.bias 1 (1280,)
  Converting to float32
model.decoder.layers.3.final_layer_norm.weight  ->  decoder.blocks.3.mlp_ln.weight
decoder.blocks.3.mlp_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.3.final_layer_norm.bias  ->  decoder.blocks.3.mlp_ln.bias
decoder.blocks.3.mlp_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.4.self_attn.k_proj.weight  ->  decoder.blocks.4.attn.key.weight
decoder.blocks.4.attn.key.weight 2 (1280, 1280)
model.decoder.layers.4.self_attn.v_proj.weight  ->  decoder.blocks.4.attn.value.weight
decoder.blocks.4.attn.value.weight 2 (1280, 1280)
model.decoder.layers.4.self_attn.v_proj.bias  ->  decoder.blocks.4.attn.value.bias
decoder.blocks.4.attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.4.self_attn.q_proj.weight  ->  decoder.blocks.4.attn.query.weight
decoder.blocks.4.attn.query.weight 2 (1280, 1280)
model.decoder.layers.4.self_attn.q_proj.bias  ->  decoder.blocks.4.attn.query.bias
decoder.blocks.4.attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.4.self_attn.out_proj.weight  ->  decoder.blocks.4.attn.out.weight
decoder.blocks.4.attn.out.weight 2 (1280, 1280)
model.decoder.layers.4.self_attn.out_proj.bias  ->  decoder.blocks.4.attn.out.bias
decoder.blocks.4.attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.4.self_attn_layer_norm.weight  ->  decoder.blocks.4.attn_ln.weight
decoder.blocks.4.attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.4.self_attn_layer_norm.bias  ->  decoder.blocks.4.attn_ln.bias
decoder.blocks.4.attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.4.encoder_attn.k_proj.weight  ->  decoder.blocks.4.cross_attn.key.weight
decoder.blocks.4.cross_attn.key.weight 2 (1280, 1280)
model.decoder.layers.4.encoder_attn.v_proj.weight  ->  decoder.blocks.4.cross_attn.value.weight
decoder.blocks.4.cross_attn.value.weight 2 (1280, 1280)
model.decoder.layers.4.encoder_attn.v_proj.bias  ->  decoder.blocks.4.cross_attn.value.bias
decoder.blocks.4.cross_attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.4.encoder_attn.q_proj.weight  ->  decoder.blocks.4.cross_attn.query.weight
decoder.blocks.4.cross_attn.query.weight 2 (1280, 1280)
model.decoder.layers.4.encoder_attn.q_proj.bias  ->  decoder.blocks.4.cross_attn.query.bias
decoder.blocks.4.cross_attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.4.encoder_attn.out_proj.weight  ->  decoder.blocks.4.cross_attn.out.weight
decoder.blocks.4.cross_attn.out.weight 2 (1280, 1280)
model.decoder.layers.4.encoder_attn.out_proj.bias  ->  decoder.blocks.4.cross_attn.out.bias
decoder.blocks.4.cross_attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.4.encoder_attn_layer_norm.weight  ->  decoder.blocks.4.cross_attn_ln.weight
decoder.blocks.4.cross_attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.4.encoder_attn_layer_norm.bias  ->  decoder.blocks.4.cross_attn_ln.bias
decoder.blocks.4.cross_attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.4.fc1.weight  ->  decoder.blocks.4.mlp.0.weight
decoder.blocks.4.mlp.0.weight 2 (5120, 1280)
model.decoder.layers.4.fc1.bias  ->  decoder.blocks.4.mlp.0.bias
decoder.blocks.4.mlp.0.bias 1 (5120,)
  Converting to float32
model.decoder.layers.4.fc2.weight  ->  decoder.blocks.4.mlp.2.weight
decoder.blocks.4.mlp.2.weight 2 (1280, 5120)
model.decoder.layers.4.fc2.bias  ->  decoder.blocks.4.mlp.2.bias
decoder.blocks.4.mlp.2.bias 1 (1280,)
  Converting to float32
model.decoder.layers.4.final_layer_norm.weight  ->  decoder.blocks.4.mlp_ln.weight
decoder.blocks.4.mlp_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.4.final_layer_norm.bias  ->  decoder.blocks.4.mlp_ln.bias
decoder.blocks.4.mlp_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.5.self_attn.k_proj.weight  ->  decoder.blocks.5.attn.key.weight
decoder.blocks.5.attn.key.weight 2 (1280, 1280)
model.decoder.layers.5.self_attn.v_proj.weight  ->  decoder.blocks.5.attn.value.weight
decoder.blocks.5.attn.value.weight 2 (1280, 1280)
model.decoder.layers.5.self_attn.v_proj.bias  ->  decoder.blocks.5.attn.value.bias
decoder.blocks.5.attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.5.self_attn.q_proj.weight  ->  decoder.blocks.5.attn.query.weight
decoder.blocks.5.attn.query.weight 2 (1280, 1280)
model.decoder.layers.5.self_attn.q_proj.bias  ->  decoder.blocks.5.attn.query.bias
decoder.blocks.5.attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.5.self_attn.out_proj.weight  ->  decoder.blocks.5.attn.out.weight
decoder.blocks.5.attn.out.weight 2 (1280, 1280)
model.decoder.layers.5.self_attn.out_proj.bias  ->  decoder.blocks.5.attn.out.bias
decoder.blocks.5.attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.5.self_attn_layer_norm.weight  ->  decoder.blocks.5.attn_ln.weight
decoder.blocks.5.attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.5.self_attn_layer_norm.bias  ->  decoder.blocks.5.attn_ln.bias
decoder.blocks.5.attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.5.encoder_attn.k_proj.weight  ->  decoder.blocks.5.cross_attn.key.weight
decoder.blocks.5.cross_attn.key.weight 2 (1280, 1280)
model.decoder.layers.5.encoder_attn.v_proj.weight  ->  decoder.blocks.5.cross_attn.value.weight
decoder.blocks.5.cross_attn.value.weight 2 (1280, 1280)
model.decoder.layers.5.encoder_attn.v_proj.bias  ->  decoder.blocks.5.cross_attn.value.bias
decoder.blocks.5.cross_attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.5.encoder_attn.q_proj.weight  ->  decoder.blocks.5.cross_attn.query.weight
decoder.blocks.5.cross_attn.query.weight 2 (1280, 1280)
model.decoder.layers.5.encoder_attn.q_proj.bias  ->  decoder.blocks.5.cross_attn.query.bias
decoder.blocks.5.cross_attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.5.encoder_attn.out_proj.weight  ->  decoder.blocks.5.cross_attn.out.weight
decoder.blocks.5.cross_attn.out.weight 2 (1280, 1280)
model.decoder.layers.5.encoder_attn.out_proj.bias  ->  decoder.blocks.5.cross_attn.out.bias
decoder.blocks.5.cross_attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.5.encoder_attn_layer_norm.weight  ->  decoder.blocks.5.cross_attn_ln.weight
decoder.blocks.5.cross_attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.5.encoder_attn_layer_norm.bias  ->  decoder.blocks.5.cross_attn_ln.bias
decoder.blocks.5.cross_attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.5.fc1.weight  ->  decoder.blocks.5.mlp.0.weight
decoder.blocks.5.mlp.0.weight 2 (5120, 1280)
model.decoder.layers.5.fc1.bias  ->  decoder.blocks.5.mlp.0.bias
decoder.blocks.5.mlp.0.bias 1 (5120,)
  Converting to float32
model.decoder.layers.5.fc2.weight  ->  decoder.blocks.5.mlp.2.weight
decoder.blocks.5.mlp.2.weight 2 (1280, 5120)
model.decoder.layers.5.fc2.bias  ->  decoder.blocks.5.mlp.2.bias
decoder.blocks.5.mlp.2.bias 1 (1280,)
  Converting to float32
model.decoder.layers.5.final_layer_norm.weight  ->  decoder.blocks.5.mlp_ln.weight
decoder.blocks.5.mlp_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.5.final_layer_norm.bias  ->  decoder.blocks.5.mlp_ln.bias
decoder.blocks.5.mlp_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.6.self_attn.k_proj.weight  ->  decoder.blocks.6.attn.key.weight
decoder.blocks.6.attn.key.weight 2 (1280, 1280)
model.decoder.layers.6.self_attn.v_proj.weight  ->  decoder.blocks.6.attn.value.weight
decoder.blocks.6.attn.value.weight 2 (1280, 1280)
model.decoder.layers.6.self_attn.v_proj.bias  ->  decoder.blocks.6.attn.value.bias
decoder.blocks.6.attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.6.self_attn.q_proj.weight  ->  decoder.blocks.6.attn.query.weight
decoder.blocks.6.attn.query.weight 2 (1280, 1280)
model.decoder.layers.6.self_attn.q_proj.bias  ->  decoder.blocks.6.attn.query.bias
decoder.blocks.6.attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.6.self_attn.out_proj.weight  ->  decoder.blocks.6.attn.out.weight
decoder.blocks.6.attn.out.weight 2 (1280, 1280)
model.decoder.layers.6.self_attn.out_proj.bias  ->  decoder.blocks.6.attn.out.bias
decoder.blocks.6.attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.6.self_attn_layer_norm.weight  ->  decoder.blocks.6.attn_ln.weight
decoder.blocks.6.attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.6.self_attn_layer_norm.bias  ->  decoder.blocks.6.attn_ln.bias
decoder.blocks.6.attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.6.encoder_attn.k_proj.weight  ->  decoder.blocks.6.cross_attn.key.weight
decoder.blocks.6.cross_attn.key.weight 2 (1280, 1280)
model.decoder.layers.6.encoder_attn.v_proj.weight  ->  decoder.blocks.6.cross_attn.value.weight
decoder.blocks.6.cross_attn.value.weight 2 (1280, 1280)
model.decoder.layers.6.encoder_attn.v_proj.bias  ->  decoder.blocks.6.cross_attn.value.bias
decoder.blocks.6.cross_attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.6.encoder_attn.q_proj.weight  ->  decoder.blocks.6.cross_attn.query.weight
decoder.blocks.6.cross_attn.query.weight 2 (1280, 1280)
model.decoder.layers.6.encoder_attn.q_proj.bias  ->  decoder.blocks.6.cross_attn.query.bias
decoder.blocks.6.cross_attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.6.encoder_attn.out_proj.weight  ->  decoder.blocks.6.cross_attn.out.weight
decoder.blocks.6.cross_attn.out.weight 2 (1280, 1280)
model.decoder.layers.6.encoder_attn.out_proj.bias  ->  decoder.blocks.6.cross_attn.out.bias
decoder.blocks.6.cross_attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.6.encoder_attn_layer_norm.weight  ->  decoder.blocks.6.cross_attn_ln.weight
decoder.blocks.6.cross_attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.6.encoder_attn_layer_norm.bias  ->  decoder.blocks.6.cross_attn_ln.bias
decoder.blocks.6.cross_attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.6.fc1.weight  ->  decoder.blocks.6.mlp.0.weight
decoder.blocks.6.mlp.0.weight 2 (5120, 1280)
model.decoder.layers.6.fc1.bias  ->  decoder.blocks.6.mlp.0.bias
decoder.blocks.6.mlp.0.bias 1 (5120,)
  Converting to float32
model.decoder.layers.6.fc2.weight  ->  decoder.blocks.6.mlp.2.weight
decoder.blocks.6.mlp.2.weight 2 (1280, 5120)
model.decoder.layers.6.fc2.bias  ->  decoder.blocks.6.mlp.2.bias
decoder.blocks.6.mlp.2.bias 1 (1280,)
  Converting to float32
model.decoder.layers.6.final_layer_norm.weight  ->  decoder.blocks.6.mlp_ln.weight
decoder.blocks.6.mlp_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.6.final_layer_norm.bias  ->  decoder.blocks.6.mlp_ln.bias
decoder.blocks.6.mlp_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.7.self_attn.k_proj.weight  ->  decoder.blocks.7.attn.key.weight
decoder.blocks.7.attn.key.weight 2 (1280, 1280)
model.decoder.layers.7.self_attn.v_proj.weight  ->  decoder.blocks.7.attn.value.weight
decoder.blocks.7.attn.value.weight 2 (1280, 1280)
model.decoder.layers.7.self_attn.v_proj.bias  ->  decoder.blocks.7.attn.value.bias
decoder.blocks.7.attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.7.self_attn.q_proj.weight  ->  decoder.blocks.7.attn.query.weight
decoder.blocks.7.attn.query.weight 2 (1280, 1280)
model.decoder.layers.7.self_attn.q_proj.bias  ->  decoder.blocks.7.attn.query.bias
decoder.blocks.7.attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.7.self_attn.out_proj.weight  ->  decoder.blocks.7.attn.out.weight
decoder.blocks.7.attn.out.weight 2 (1280, 1280)
model.decoder.layers.7.self_attn.out_proj.bias  ->  decoder.blocks.7.attn.out.bias
decoder.blocks.7.attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.7.self_attn_layer_norm.weight  ->  decoder.blocks.7.attn_ln.weight
decoder.blocks.7.attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.7.self_attn_layer_norm.bias  ->  decoder.blocks.7.attn_ln.bias
decoder.blocks.7.attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.7.encoder_attn.k_proj.weight  ->  decoder.blocks.7.cross_attn.key.weight
decoder.blocks.7.cross_attn.key.weight 2 (1280, 1280)
model.decoder.layers.7.encoder_attn.v_proj.weight  ->  decoder.blocks.7.cross_attn.value.weight
decoder.blocks.7.cross_attn.value.weight 2 (1280, 1280)
model.decoder.layers.7.encoder_attn.v_proj.bias  ->  decoder.blocks.7.cross_attn.value.bias
decoder.blocks.7.cross_attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.7.encoder_attn.q_proj.weight  ->  decoder.blocks.7.cross_attn.query.weight
decoder.blocks.7.cross_attn.query.weight 2 (1280, 1280)
model.decoder.layers.7.encoder_attn.q_proj.bias  ->  decoder.blocks.7.cross_attn.query.bias
decoder.blocks.7.cross_attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.7.encoder_attn.out_proj.weight  ->  decoder.blocks.7.cross_attn.out.weight
decoder.blocks.7.cross_attn.out.weight 2 (1280, 1280)
model.decoder.layers.7.encoder_attn.out_proj.bias  ->  decoder.blocks.7.cross_attn.out.bias
decoder.blocks.7.cross_attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.7.encoder_attn_layer_norm.weight  ->  decoder.blocks.7.cross_attn_ln.weight
decoder.blocks.7.cross_attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.7.encoder_attn_layer_norm.bias  ->  decoder.blocks.7.cross_attn_ln.bias
decoder.blocks.7.cross_attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.7.fc1.weight  ->  decoder.blocks.7.mlp.0.weight
decoder.blocks.7.mlp.0.weight 2 (5120, 1280)
model.decoder.layers.7.fc1.bias  ->  decoder.blocks.7.mlp.0.bias
decoder.blocks.7.mlp.0.bias 1 (5120,)
  Converting to float32
model.decoder.layers.7.fc2.weight  ->  decoder.blocks.7.mlp.2.weight
decoder.blocks.7.mlp.2.weight 2 (1280, 5120)
model.decoder.layers.7.fc2.bias  ->  decoder.blocks.7.mlp.2.bias
decoder.blocks.7.mlp.2.bias 1 (1280,)
  Converting to float32
model.decoder.layers.7.final_layer_norm.weight  ->  decoder.blocks.7.mlp_ln.weight
decoder.blocks.7.mlp_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.7.final_layer_norm.bias  ->  decoder.blocks.7.mlp_ln.bias
decoder.blocks.7.mlp_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.8.self_attn.k_proj.weight  ->  decoder.blocks.8.attn.key.weight
decoder.blocks.8.attn.key.weight 2 (1280, 1280)
model.decoder.layers.8.self_attn.v_proj.weight  ->  decoder.blocks.8.attn.value.weight
decoder.blocks.8.attn.value.weight 2 (1280, 1280)
model.decoder.layers.8.self_attn.v_proj.bias  ->  decoder.blocks.8.attn.value.bias
decoder.blocks.8.attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.8.self_attn.q_proj.weight  ->  decoder.blocks.8.attn.query.weight
decoder.blocks.8.attn.query.weight 2 (1280, 1280)
model.decoder.layers.8.self_attn.q_proj.bias  ->  decoder.blocks.8.attn.query.bias
decoder.blocks.8.attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.8.self_attn.out_proj.weight  ->  decoder.blocks.8.attn.out.weight
decoder.blocks.8.attn.out.weight 2 (1280, 1280)
model.decoder.layers.8.self_attn.out_proj.bias  ->  decoder.blocks.8.attn.out.bias
decoder.blocks.8.attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.8.self_attn_layer_norm.weight  ->  decoder.blocks.8.attn_ln.weight
decoder.blocks.8.attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.8.self_attn_layer_norm.bias  ->  decoder.blocks.8.attn_ln.bias
decoder.blocks.8.attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.8.encoder_attn.k_proj.weight  ->  decoder.blocks.8.cross_attn.key.weight
decoder.blocks.8.cross_attn.key.weight 2 (1280, 1280)
model.decoder.layers.8.encoder_attn.v_proj.weight  ->  decoder.blocks.8.cross_attn.value.weight
decoder.blocks.8.cross_attn.value.weight 2 (1280, 1280)
model.decoder.layers.8.encoder_attn.v_proj.bias  ->  decoder.blocks.8.cross_attn.value.bias
decoder.blocks.8.cross_attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.8.encoder_attn.q_proj.weight  ->  decoder.blocks.8.cross_attn.query.weight
decoder.blocks.8.cross_attn.query.weight 2 (1280, 1280)
model.decoder.layers.8.encoder_attn.q_proj.bias  ->  decoder.blocks.8.cross_attn.query.bias
decoder.blocks.8.cross_attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.8.encoder_attn.out_proj.weight  ->  decoder.blocks.8.cross_attn.out.weight
decoder.blocks.8.cross_attn.out.weight 2 (1280, 1280)
model.decoder.layers.8.encoder_attn.out_proj.bias  ->  decoder.blocks.8.cross_attn.out.bias
decoder.blocks.8.cross_attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.8.encoder_attn_layer_norm.weight  ->  decoder.blocks.8.cross_attn_ln.weight
decoder.blocks.8.cross_attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.8.encoder_attn_layer_norm.bias  ->  decoder.blocks.8.cross_attn_ln.bias
decoder.blocks.8.cross_attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.8.fc1.weight  ->  decoder.blocks.8.mlp.0.weight
decoder.blocks.8.mlp.0.weight 2 (5120, 1280)
model.decoder.layers.8.fc1.bias  ->  decoder.blocks.8.mlp.0.bias
decoder.blocks.8.mlp.0.bias 1 (5120,)
  Converting to float32
model.decoder.layers.8.fc2.weight  ->  decoder.blocks.8.mlp.2.weight
decoder.blocks.8.mlp.2.weight 2 (1280, 5120)
model.decoder.layers.8.fc2.bias  ->  decoder.blocks.8.mlp.2.bias
decoder.blocks.8.mlp.2.bias 1 (1280,)
  Converting to float32
model.decoder.layers.8.final_layer_norm.weight  ->  decoder.blocks.8.mlp_ln.weight
decoder.blocks.8.mlp_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.8.final_layer_norm.bias  ->  decoder.blocks.8.mlp_ln.bias
decoder.blocks.8.mlp_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.9.self_attn.k_proj.weight  ->  decoder.blocks.9.attn.key.weight
decoder.blocks.9.attn.key.weight 2 (1280, 1280)
model.decoder.layers.9.self_attn.v_proj.weight  ->  decoder.blocks.9.attn.value.weight
decoder.blocks.9.attn.value.weight 2 (1280, 1280)
model.decoder.layers.9.self_attn.v_proj.bias  ->  decoder.blocks.9.attn.value.bias
decoder.blocks.9.attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.9.self_attn.q_proj.weight  ->  decoder.blocks.9.attn.query.weight
decoder.blocks.9.attn.query.weight 2 (1280, 1280)
model.decoder.layers.9.self_attn.q_proj.bias  ->  decoder.blocks.9.attn.query.bias
decoder.blocks.9.attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.9.self_attn.out_proj.weight  ->  decoder.blocks.9.attn.out.weight
decoder.blocks.9.attn.out.weight 2 (1280, 1280)
model.decoder.layers.9.self_attn.out_proj.bias  ->  decoder.blocks.9.attn.out.bias
decoder.blocks.9.attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.9.self_attn_layer_norm.weight  ->  decoder.blocks.9.attn_ln.weight
decoder.blocks.9.attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.9.self_attn_layer_norm.bias  ->  decoder.blocks.9.attn_ln.bias
decoder.blocks.9.attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.9.encoder_attn.k_proj.weight  ->  decoder.blocks.9.cross_attn.key.weight
decoder.blocks.9.cross_attn.key.weight 2 (1280, 1280)
model.decoder.layers.9.encoder_attn.v_proj.weight  ->  decoder.blocks.9.cross_attn.value.weight
decoder.blocks.9.cross_attn.value.weight 2 (1280, 1280)
model.decoder.layers.9.encoder_attn.v_proj.bias  ->  decoder.blocks.9.cross_attn.value.bias
decoder.blocks.9.cross_attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.9.encoder_attn.q_proj.weight  ->  decoder.blocks.9.cross_attn.query.weight
decoder.blocks.9.cross_attn.query.weight 2 (1280, 1280)
model.decoder.layers.9.encoder_attn.q_proj.bias  ->  decoder.blocks.9.cross_attn.query.bias
decoder.blocks.9.cross_attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.9.encoder_attn.out_proj.weight  ->  decoder.blocks.9.cross_attn.out.weight
decoder.blocks.9.cross_attn.out.weight 2 (1280, 1280)
model.decoder.layers.9.encoder_attn.out_proj.bias  ->  decoder.blocks.9.cross_attn.out.bias
decoder.blocks.9.cross_attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.9.encoder_attn_layer_norm.weight  ->  decoder.blocks.9.cross_attn_ln.weight
decoder.blocks.9.cross_attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.9.encoder_attn_layer_norm.bias  ->  decoder.blocks.9.cross_attn_ln.bias
decoder.blocks.9.cross_attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.9.fc1.weight  ->  decoder.blocks.9.mlp.0.weight
decoder.blocks.9.mlp.0.weight 2 (5120, 1280)
model.decoder.layers.9.fc1.bias  ->  decoder.blocks.9.mlp.0.bias
decoder.blocks.9.mlp.0.bias 1 (5120,)
  Converting to float32
model.decoder.layers.9.fc2.weight  ->  decoder.blocks.9.mlp.2.weight
decoder.blocks.9.mlp.2.weight 2 (1280, 5120)
model.decoder.layers.9.fc2.bias  ->  decoder.blocks.9.mlp.2.bias
decoder.blocks.9.mlp.2.bias 1 (1280,)
  Converting to float32
model.decoder.layers.9.final_layer_norm.weight  ->  decoder.blocks.9.mlp_ln.weight
decoder.blocks.9.mlp_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.9.final_layer_norm.bias  ->  decoder.blocks.9.mlp_ln.bias
decoder.blocks.9.mlp_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.10.self_attn.k_proj.weight  ->  decoder.blocks.10.attn.key.weight
decoder.blocks.10.attn.key.weight 2 (1280, 1280)
model.decoder.layers.10.self_attn.v_proj.weight  ->  decoder.blocks.10.attn.value.weight
decoder.blocks.10.attn.value.weight 2 (1280, 1280)
model.decoder.layers.10.self_attn.v_proj.bias  ->  decoder.blocks.10.attn.value.bias
decoder.blocks.10.attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.10.self_attn.q_proj.weight  ->  decoder.blocks.10.attn.query.weight
decoder.blocks.10.attn.query.weight 2 (1280, 1280)
model.decoder.layers.10.self_attn.q_proj.bias  ->  decoder.blocks.10.attn.query.bias
decoder.blocks.10.attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.10.self_attn.out_proj.weight  ->  decoder.blocks.10.attn.out.weight
decoder.blocks.10.attn.out.weight 2 (1280, 1280)
model.decoder.layers.10.self_attn.out_proj.bias  ->  decoder.blocks.10.attn.out.bias
decoder.blocks.10.attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.10.self_attn_layer_norm.weight  ->  decoder.blocks.10.attn_ln.weight
decoder.blocks.10.attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.10.self_attn_layer_norm.bias  ->  decoder.blocks.10.attn_ln.bias
decoder.blocks.10.attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.10.encoder_attn.k_proj.weight  ->  decoder.blocks.10.cross_attn.key.weight
decoder.blocks.10.cross_attn.key.weight 2 (1280, 1280)
model.decoder.layers.10.encoder_attn.v_proj.weight  ->  decoder.blocks.10.cross_attn.value.weight
decoder.blocks.10.cross_attn.value.weight 2 (1280, 1280)
model.decoder.layers.10.encoder_attn.v_proj.bias  ->  decoder.blocks.10.cross_attn.value.bias
decoder.blocks.10.cross_attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.10.encoder_attn.q_proj.weight  ->  decoder.blocks.10.cross_attn.query.weight
decoder.blocks.10.cross_attn.query.weight 2 (1280, 1280)
model.decoder.layers.10.encoder_attn.q_proj.bias  ->  decoder.blocks.10.cross_attn.query.bias
decoder.blocks.10.cross_attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.10.encoder_attn.out_proj.weight  ->  decoder.blocks.10.cross_attn.out.weight
decoder.blocks.10.cross_attn.out.weight 2 (1280, 1280)
model.decoder.layers.10.encoder_attn.out_proj.bias  ->  decoder.blocks.10.cross_attn.out.bias
decoder.blocks.10.cross_attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.10.encoder_attn_layer_norm.weight  ->  decoder.blocks.10.cross_attn_ln.weight
decoder.blocks.10.cross_attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.10.encoder_attn_layer_norm.bias  ->  decoder.blocks.10.cross_attn_ln.bias
decoder.blocks.10.cross_attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.10.fc1.weight  ->  decoder.blocks.10.mlp.0.weight
decoder.blocks.10.mlp.0.weight 2 (5120, 1280)
model.decoder.layers.10.fc1.bias  ->  decoder.blocks.10.mlp.0.bias
decoder.blocks.10.mlp.0.bias 1 (5120,)
  Converting to float32
model.decoder.layers.10.fc2.weight  ->  decoder.blocks.10.mlp.2.weight
decoder.blocks.10.mlp.2.weight 2 (1280, 5120)
model.decoder.layers.10.fc2.bias  ->  decoder.blocks.10.mlp.2.bias
decoder.blocks.10.mlp.2.bias 1 (1280,)
  Converting to float32
model.decoder.layers.10.final_layer_norm.weight  ->  decoder.blocks.10.mlp_ln.weight
decoder.blocks.10.mlp_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.10.final_layer_norm.bias  ->  decoder.blocks.10.mlp_ln.bias
decoder.blocks.10.mlp_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.11.self_attn.k_proj.weight  ->  decoder.blocks.11.attn.key.weight
decoder.blocks.11.attn.key.weight 2 (1280, 1280)
model.decoder.layers.11.self_attn.v_proj.weight  ->  decoder.blocks.11.attn.value.weight
decoder.blocks.11.attn.value.weight 2 (1280, 1280)
model.decoder.layers.11.self_attn.v_proj.bias  ->  decoder.blocks.11.attn.value.bias
decoder.blocks.11.attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.11.self_attn.q_proj.weight  ->  decoder.blocks.11.attn.query.weight
decoder.blocks.11.attn.query.weight 2 (1280, 1280)
model.decoder.layers.11.self_attn.q_proj.bias  ->  decoder.blocks.11.attn.query.bias
decoder.blocks.11.attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.11.self_attn.out_proj.weight  ->  decoder.blocks.11.attn.out.weight
decoder.blocks.11.attn.out.weight 2 (1280, 1280)
model.decoder.layers.11.self_attn.out_proj.bias  ->  decoder.blocks.11.attn.out.bias
decoder.blocks.11.attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.11.self_attn_layer_norm.weight  ->  decoder.blocks.11.attn_ln.weight
decoder.blocks.11.attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.11.self_attn_layer_norm.bias  ->  decoder.blocks.11.attn_ln.bias
decoder.blocks.11.attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.11.encoder_attn.k_proj.weight  ->  decoder.blocks.11.cross_attn.key.weight
decoder.blocks.11.cross_attn.key.weight 2 (1280, 1280)
model.decoder.layers.11.encoder_attn.v_proj.weight  ->  decoder.blocks.11.cross_attn.value.weight
decoder.blocks.11.cross_attn.value.weight 2 (1280, 1280)
model.decoder.layers.11.encoder_attn.v_proj.bias  ->  decoder.blocks.11.cross_attn.value.bias
decoder.blocks.11.cross_attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.11.encoder_attn.q_proj.weight  ->  decoder.blocks.11.cross_attn.query.weight
decoder.blocks.11.cross_attn.query.weight 2 (1280, 1280)
model.decoder.layers.11.encoder_attn.q_proj.bias  ->  decoder.blocks.11.cross_attn.query.bias
decoder.blocks.11.cross_attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.11.encoder_attn.out_proj.weight  ->  decoder.blocks.11.cross_attn.out.weight
decoder.blocks.11.cross_attn.out.weight 2 (1280, 1280)
model.decoder.layers.11.encoder_attn.out_proj.bias  ->  decoder.blocks.11.cross_attn.out.bias
decoder.blocks.11.cross_attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.11.encoder_attn_layer_norm.weight  ->  decoder.blocks.11.cross_attn_ln.weight
decoder.blocks.11.cross_attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.11.encoder_attn_layer_norm.bias  ->  decoder.blocks.11.cross_attn_ln.bias
decoder.blocks.11.cross_attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.11.fc1.weight  ->  decoder.blocks.11.mlp.0.weight
decoder.blocks.11.mlp.0.weight 2 (5120, 1280)
model.decoder.layers.11.fc1.bias  ->  decoder.blocks.11.mlp.0.bias
decoder.blocks.11.mlp.0.bias 1 (5120,)
  Converting to float32
model.decoder.layers.11.fc2.weight  ->  decoder.blocks.11.mlp.2.weight
decoder.blocks.11.mlp.2.weight 2 (1280, 5120)
model.decoder.layers.11.fc2.bias  ->  decoder.blocks.11.mlp.2.bias
decoder.blocks.11.mlp.2.bias 1 (1280,)
  Converting to float32
model.decoder.layers.11.final_layer_norm.weight  ->  decoder.blocks.11.mlp_ln.weight
decoder.blocks.11.mlp_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.11.final_layer_norm.bias  ->  decoder.blocks.11.mlp_ln.bias
decoder.blocks.11.mlp_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.12.self_attn.k_proj.weight  ->  decoder.blocks.12.attn.key.weight
decoder.blocks.12.attn.key.weight 2 (1280, 1280)
model.decoder.layers.12.self_attn.v_proj.weight  ->  decoder.blocks.12.attn.value.weight
decoder.blocks.12.attn.value.weight 2 (1280, 1280)
model.decoder.layers.12.self_attn.v_proj.bias  ->  decoder.blocks.12.attn.value.bias
decoder.blocks.12.attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.12.self_attn.q_proj.weight  ->  decoder.blocks.12.attn.query.weight
decoder.blocks.12.attn.query.weight 2 (1280, 1280)
model.decoder.layers.12.self_attn.q_proj.bias  ->  decoder.blocks.12.attn.query.bias
decoder.blocks.12.attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.12.self_attn.out_proj.weight  ->  decoder.blocks.12.attn.out.weight
decoder.blocks.12.attn.out.weight 2 (1280, 1280)
model.decoder.layers.12.self_attn.out_proj.bias  ->  decoder.blocks.12.attn.out.bias
decoder.blocks.12.attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.12.self_attn_layer_norm.weight  ->  decoder.blocks.12.attn_ln.weight
decoder.blocks.12.attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.12.self_attn_layer_norm.bias  ->  decoder.blocks.12.attn_ln.bias
decoder.blocks.12.attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.12.encoder_attn.k_proj.weight  ->  decoder.blocks.12.cross_attn.key.weight
decoder.blocks.12.cross_attn.key.weight 2 (1280, 1280)
model.decoder.layers.12.encoder_attn.v_proj.weight  ->  decoder.blocks.12.cross_attn.value.weight
decoder.blocks.12.cross_attn.value.weight 2 (1280, 1280)
model.decoder.layers.12.encoder_attn.v_proj.bias  ->  decoder.blocks.12.cross_attn.value.bias
decoder.blocks.12.cross_attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.12.encoder_attn.q_proj.weight  ->  decoder.blocks.12.cross_attn.query.weight
decoder.blocks.12.cross_attn.query.weight 2 (1280, 1280)
model.decoder.layers.12.encoder_attn.q_proj.bias  ->  decoder.blocks.12.cross_attn.query.bias
decoder.blocks.12.cross_attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.12.encoder_attn.out_proj.weight  ->  decoder.blocks.12.cross_attn.out.weight
decoder.blocks.12.cross_attn.out.weight 2 (1280, 1280)
model.decoder.layers.12.encoder_attn.out_proj.bias  ->  decoder.blocks.12.cross_attn.out.bias
decoder.blocks.12.cross_attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.12.encoder_attn_layer_norm.weight  ->  decoder.blocks.12.cross_attn_ln.weight
decoder.blocks.12.cross_attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.12.encoder_attn_layer_norm.bias  ->  decoder.blocks.12.cross_attn_ln.bias
decoder.blocks.12.cross_attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.12.fc1.weight  ->  decoder.blocks.12.mlp.0.weight
decoder.blocks.12.mlp.0.weight 2 (5120, 1280)
model.decoder.layers.12.fc1.bias  ->  decoder.blocks.12.mlp.0.bias
decoder.blocks.12.mlp.0.bias 1 (5120,)
  Converting to float32
model.decoder.layers.12.fc2.weight  ->  decoder.blocks.12.mlp.2.weight
decoder.blocks.12.mlp.2.weight 2 (1280, 5120)
model.decoder.layers.12.fc2.bias  ->  decoder.blocks.12.mlp.2.bias
decoder.blocks.12.mlp.2.bias 1 (1280,)
  Converting to float32
model.decoder.layers.12.final_layer_norm.weight  ->  decoder.blocks.12.mlp_ln.weight
decoder.blocks.12.mlp_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.12.final_layer_norm.bias  ->  decoder.blocks.12.mlp_ln.bias
decoder.blocks.12.mlp_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.13.self_attn.k_proj.weight  ->  decoder.blocks.13.attn.key.weight
decoder.blocks.13.attn.key.weight 2 (1280, 1280)
model.decoder.layers.13.self_attn.v_proj.weight  ->  decoder.blocks.13.attn.value.weight
decoder.blocks.13.attn.value.weight 2 (1280, 1280)
model.decoder.layers.13.self_attn.v_proj.bias  ->  decoder.blocks.13.attn.value.bias
decoder.blocks.13.attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.13.self_attn.q_proj.weight  ->  decoder.blocks.13.attn.query.weight
decoder.blocks.13.attn.query.weight 2 (1280, 1280)
model.decoder.layers.13.self_attn.q_proj.bias  ->  decoder.blocks.13.attn.query.bias
decoder.blocks.13.attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.13.self_attn.out_proj.weight  ->  decoder.blocks.13.attn.out.weight
decoder.blocks.13.attn.out.weight 2 (1280, 1280)
model.decoder.layers.13.self_attn.out_proj.bias  ->  decoder.blocks.13.attn.out.bias
decoder.blocks.13.attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.13.self_attn_layer_norm.weight  ->  decoder.blocks.13.attn_ln.weight
decoder.blocks.13.attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.13.self_attn_layer_norm.bias  ->  decoder.blocks.13.attn_ln.bias
decoder.blocks.13.attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.13.encoder_attn.k_proj.weight  ->  decoder.blocks.13.cross_attn.key.weight
decoder.blocks.13.cross_attn.key.weight 2 (1280, 1280)
model.decoder.layers.13.encoder_attn.v_proj.weight  ->  decoder.blocks.13.cross_attn.value.weight
decoder.blocks.13.cross_attn.value.weight 2 (1280, 1280)
model.decoder.layers.13.encoder_attn.v_proj.bias  ->  decoder.blocks.13.cross_attn.value.bias
decoder.blocks.13.cross_attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.13.encoder_attn.q_proj.weight  ->  decoder.blocks.13.cross_attn.query.weight
decoder.blocks.13.cross_attn.query.weight 2 (1280, 1280)
model.decoder.layers.13.encoder_attn.q_proj.bias  ->  decoder.blocks.13.cross_attn.query.bias
decoder.blocks.13.cross_attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.13.encoder_attn.out_proj.weight  ->  decoder.blocks.13.cross_attn.out.weight
decoder.blocks.13.cross_attn.out.weight 2 (1280, 1280)
model.decoder.layers.13.encoder_attn.out_proj.bias  ->  decoder.blocks.13.cross_attn.out.bias
decoder.blocks.13.cross_attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.13.encoder_attn_layer_norm.weight  ->  decoder.blocks.13.cross_attn_ln.weight
decoder.blocks.13.cross_attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.13.encoder_attn_layer_norm.bias  ->  decoder.blocks.13.cross_attn_ln.bias
decoder.blocks.13.cross_attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.13.fc1.weight  ->  decoder.blocks.13.mlp.0.weight
decoder.blocks.13.mlp.0.weight 2 (5120, 1280)
model.decoder.layers.13.fc1.bias  ->  decoder.blocks.13.mlp.0.bias
decoder.blocks.13.mlp.0.bias 1 (5120,)
  Converting to float32
model.decoder.layers.13.fc2.weight  ->  decoder.blocks.13.mlp.2.weight
decoder.blocks.13.mlp.2.weight 2 (1280, 5120)
model.decoder.layers.13.fc2.bias  ->  decoder.blocks.13.mlp.2.bias
decoder.blocks.13.mlp.2.bias 1 (1280,)
  Converting to float32
model.decoder.layers.13.final_layer_norm.weight  ->  decoder.blocks.13.mlp_ln.weight
decoder.blocks.13.mlp_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.13.final_layer_norm.bias  ->  decoder.blocks.13.mlp_ln.bias
decoder.blocks.13.mlp_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.14.self_attn.k_proj.weight  ->  decoder.blocks.14.attn.key.weight
decoder.blocks.14.attn.key.weight 2 (1280, 1280)
model.decoder.layers.14.self_attn.v_proj.weight  ->  decoder.blocks.14.attn.value.weight
decoder.blocks.14.attn.value.weight 2 (1280, 1280)
model.decoder.layers.14.self_attn.v_proj.bias  ->  decoder.blocks.14.attn.value.bias
decoder.blocks.14.attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.14.self_attn.q_proj.weight  ->  decoder.blocks.14.attn.query.weight
decoder.blocks.14.attn.query.weight 2 (1280, 1280)
model.decoder.layers.14.self_attn.q_proj.bias  ->  decoder.blocks.14.attn.query.bias
decoder.blocks.14.attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.14.self_attn.out_proj.weight  ->  decoder.blocks.14.attn.out.weight
decoder.blocks.14.attn.out.weight 2 (1280, 1280)
model.decoder.layers.14.self_attn.out_proj.bias  ->  decoder.blocks.14.attn.out.bias
decoder.blocks.14.attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.14.self_attn_layer_norm.weight  ->  decoder.blocks.14.attn_ln.weight
decoder.blocks.14.attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.14.self_attn_layer_norm.bias  ->  decoder.blocks.14.attn_ln.bias
decoder.blocks.14.attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.14.encoder_attn.k_proj.weight  ->  decoder.blocks.14.cross_attn.key.weight
decoder.blocks.14.cross_attn.key.weight 2 (1280, 1280)
model.decoder.layers.14.encoder_attn.v_proj.weight  ->  decoder.blocks.14.cross_attn.value.weight
decoder.blocks.14.cross_attn.value.weight 2 (1280, 1280)
model.decoder.layers.14.encoder_attn.v_proj.bias  ->  decoder.blocks.14.cross_attn.value.bias
decoder.blocks.14.cross_attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.14.encoder_attn.q_proj.weight  ->  decoder.blocks.14.cross_attn.query.weight
decoder.blocks.14.cross_attn.query.weight 2 (1280, 1280)
model.decoder.layers.14.encoder_attn.q_proj.bias  ->  decoder.blocks.14.cross_attn.query.bias
decoder.blocks.14.cross_attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.14.encoder_attn.out_proj.weight  ->  decoder.blocks.14.cross_attn.out.weight
decoder.blocks.14.cross_attn.out.weight 2 (1280, 1280)
model.decoder.layers.14.encoder_attn.out_proj.bias  ->  decoder.blocks.14.cross_attn.out.bias
decoder.blocks.14.cross_attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.14.encoder_attn_layer_norm.weight  ->  decoder.blocks.14.cross_attn_ln.weight
decoder.blocks.14.cross_attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.14.encoder_attn_layer_norm.bias  ->  decoder.blocks.14.cross_attn_ln.bias
decoder.blocks.14.cross_attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.14.fc1.weight  ->  decoder.blocks.14.mlp.0.weight
decoder.blocks.14.mlp.0.weight 2 (5120, 1280)
model.decoder.layers.14.fc1.bias  ->  decoder.blocks.14.mlp.0.bias
decoder.blocks.14.mlp.0.bias 1 (5120,)
  Converting to float32
model.decoder.layers.14.fc2.weight  ->  decoder.blocks.14.mlp.2.weight
decoder.blocks.14.mlp.2.weight 2 (1280, 5120)
model.decoder.layers.14.fc2.bias  ->  decoder.blocks.14.mlp.2.bias
decoder.blocks.14.mlp.2.bias 1 (1280,)
  Converting to float32
model.decoder.layers.14.final_layer_norm.weight  ->  decoder.blocks.14.mlp_ln.weight
decoder.blocks.14.mlp_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.14.final_layer_norm.bias  ->  decoder.blocks.14.mlp_ln.bias
decoder.blocks.14.mlp_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.15.self_attn.k_proj.weight  ->  decoder.blocks.15.attn.key.weight
decoder.blocks.15.attn.key.weight 2 (1280, 1280)
model.decoder.layers.15.self_attn.v_proj.weight  ->  decoder.blocks.15.attn.value.weight
decoder.blocks.15.attn.value.weight 2 (1280, 1280)
model.decoder.layers.15.self_attn.v_proj.bias  ->  decoder.blocks.15.attn.value.bias
decoder.blocks.15.attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.15.self_attn.q_proj.weight  ->  decoder.blocks.15.attn.query.weight
decoder.blocks.15.attn.query.weight 2 (1280, 1280)
model.decoder.layers.15.self_attn.q_proj.bias  ->  decoder.blocks.15.attn.query.bias
decoder.blocks.15.attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.15.self_attn.out_proj.weight  ->  decoder.blocks.15.attn.out.weight
decoder.blocks.15.attn.out.weight 2 (1280, 1280)
model.decoder.layers.15.self_attn.out_proj.bias  ->  decoder.blocks.15.attn.out.bias
decoder.blocks.15.attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.15.self_attn_layer_norm.weight  ->  decoder.blocks.15.attn_ln.weight
decoder.blocks.15.attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.15.self_attn_layer_norm.bias  ->  decoder.blocks.15.attn_ln.bias
decoder.blocks.15.attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.15.encoder_attn.k_proj.weight  ->  decoder.blocks.15.cross_attn.key.weight
decoder.blocks.15.cross_attn.key.weight 2 (1280, 1280)
model.decoder.layers.15.encoder_attn.v_proj.weight  ->  decoder.blocks.15.cross_attn.value.weight
decoder.blocks.15.cross_attn.value.weight 2 (1280, 1280)
model.decoder.layers.15.encoder_attn.v_proj.bias  ->  decoder.blocks.15.cross_attn.value.bias
decoder.blocks.15.cross_attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.15.encoder_attn.q_proj.weight  ->  decoder.blocks.15.cross_attn.query.weight
decoder.blocks.15.cross_attn.query.weight 2 (1280, 1280)
model.decoder.layers.15.encoder_attn.q_proj.bias  ->  decoder.blocks.15.cross_attn.query.bias
decoder.blocks.15.cross_attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.15.encoder_attn.out_proj.weight  ->  decoder.blocks.15.cross_attn.out.weight
decoder.blocks.15.cross_attn.out.weight 2 (1280, 1280)
model.decoder.layers.15.encoder_attn.out_proj.bias  ->  decoder.blocks.15.cross_attn.out.bias
decoder.blocks.15.cross_attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.15.encoder_attn_layer_norm.weight  ->  decoder.blocks.15.cross_attn_ln.weight
decoder.blocks.15.cross_attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.15.encoder_attn_layer_norm.bias  ->  decoder.blocks.15.cross_attn_ln.bias
decoder.blocks.15.cross_attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.15.fc1.weight  ->  decoder.blocks.15.mlp.0.weight
decoder.blocks.15.mlp.0.weight 2 (5120, 1280)
model.decoder.layers.15.fc1.bias  ->  decoder.blocks.15.mlp.0.bias
decoder.blocks.15.mlp.0.bias 1 (5120,)
  Converting to float32
model.decoder.layers.15.fc2.weight  ->  decoder.blocks.15.mlp.2.weight
decoder.blocks.15.mlp.2.weight 2 (1280, 5120)
model.decoder.layers.15.fc2.bias  ->  decoder.blocks.15.mlp.2.bias
decoder.blocks.15.mlp.2.bias 1 (1280,)
  Converting to float32
model.decoder.layers.15.final_layer_norm.weight  ->  decoder.blocks.15.mlp_ln.weight
decoder.blocks.15.mlp_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.15.final_layer_norm.bias  ->  decoder.blocks.15.mlp_ln.bias
decoder.blocks.15.mlp_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.16.self_attn.k_proj.weight  ->  decoder.blocks.16.attn.key.weight
decoder.blocks.16.attn.key.weight 2 (1280, 1280)
model.decoder.layers.16.self_attn.v_proj.weight  ->  decoder.blocks.16.attn.value.weight
decoder.blocks.16.attn.value.weight 2 (1280, 1280)
model.decoder.layers.16.self_attn.v_proj.bias  ->  decoder.blocks.16.attn.value.bias
decoder.blocks.16.attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.16.self_attn.q_proj.weight  ->  decoder.blocks.16.attn.query.weight
decoder.blocks.16.attn.query.weight 2 (1280, 1280)
model.decoder.layers.16.self_attn.q_proj.bias  ->  decoder.blocks.16.attn.query.bias
decoder.blocks.16.attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.16.self_attn.out_proj.weight  ->  decoder.blocks.16.attn.out.weight
decoder.blocks.16.attn.out.weight 2 (1280, 1280)
model.decoder.layers.16.self_attn.out_proj.bias  ->  decoder.blocks.16.attn.out.bias
decoder.blocks.16.attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.16.self_attn_layer_norm.weight  ->  decoder.blocks.16.attn_ln.weight
decoder.blocks.16.attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.16.self_attn_layer_norm.bias  ->  decoder.blocks.16.attn_ln.bias
decoder.blocks.16.attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.16.encoder_attn.k_proj.weight  ->  decoder.blocks.16.cross_attn.key.weight
decoder.blocks.16.cross_attn.key.weight 2 (1280, 1280)
model.decoder.layers.16.encoder_attn.v_proj.weight  ->  decoder.blocks.16.cross_attn.value.weight
decoder.blocks.16.cross_attn.value.weight 2 (1280, 1280)
model.decoder.layers.16.encoder_attn.v_proj.bias  ->  decoder.blocks.16.cross_attn.value.bias
decoder.blocks.16.cross_attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.16.encoder_attn.q_proj.weight  ->  decoder.blocks.16.cross_attn.query.weight
decoder.blocks.16.cross_attn.query.weight 2 (1280, 1280)
model.decoder.layers.16.encoder_attn.q_proj.bias  ->  decoder.blocks.16.cross_attn.query.bias
decoder.blocks.16.cross_attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.16.encoder_attn.out_proj.weight  ->  decoder.blocks.16.cross_attn.out.weight
decoder.blocks.16.cross_attn.out.weight 2 (1280, 1280)
model.decoder.layers.16.encoder_attn.out_proj.bias  ->  decoder.blocks.16.cross_attn.out.bias
decoder.blocks.16.cross_attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.16.encoder_attn_layer_norm.weight  ->  decoder.blocks.16.cross_attn_ln.weight
decoder.blocks.16.cross_attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.16.encoder_attn_layer_norm.bias  ->  decoder.blocks.16.cross_attn_ln.bias
decoder.blocks.16.cross_attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.16.fc1.weight  ->  decoder.blocks.16.mlp.0.weight
decoder.blocks.16.mlp.0.weight 2 (5120, 1280)
model.decoder.layers.16.fc1.bias  ->  decoder.blocks.16.mlp.0.bias
decoder.blocks.16.mlp.0.bias 1 (5120,)
  Converting to float32
model.decoder.layers.16.fc2.weight  ->  decoder.blocks.16.mlp.2.weight
decoder.blocks.16.mlp.2.weight 2 (1280, 5120)
model.decoder.layers.16.fc2.bias  ->  decoder.blocks.16.mlp.2.bias
decoder.blocks.16.mlp.2.bias 1 (1280,)
  Converting to float32
model.decoder.layers.16.final_layer_norm.weight  ->  decoder.blocks.16.mlp_ln.weight
decoder.blocks.16.mlp_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.16.final_layer_norm.bias  ->  decoder.blocks.16.mlp_ln.bias
decoder.blocks.16.mlp_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.17.self_attn.k_proj.weight  ->  decoder.blocks.17.attn.key.weight
decoder.blocks.17.attn.key.weight 2 (1280, 1280)
model.decoder.layers.17.self_attn.v_proj.weight  ->  decoder.blocks.17.attn.value.weight
decoder.blocks.17.attn.value.weight 2 (1280, 1280)
model.decoder.layers.17.self_attn.v_proj.bias  ->  decoder.blocks.17.attn.value.bias
decoder.blocks.17.attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.17.self_attn.q_proj.weight  ->  decoder.blocks.17.attn.query.weight
decoder.blocks.17.attn.query.weight 2 (1280, 1280)
model.decoder.layers.17.self_attn.q_proj.bias  ->  decoder.blocks.17.attn.query.bias
decoder.blocks.17.attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.17.self_attn.out_proj.weight  ->  decoder.blocks.17.attn.out.weight
decoder.blocks.17.attn.out.weight 2 (1280, 1280)
model.decoder.layers.17.self_attn.out_proj.bias  ->  decoder.blocks.17.attn.out.bias
decoder.blocks.17.attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.17.self_attn_layer_norm.weight  ->  decoder.blocks.17.attn_ln.weight
decoder.blocks.17.attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.17.self_attn_layer_norm.bias  ->  decoder.blocks.17.attn_ln.bias
decoder.blocks.17.attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.17.encoder_attn.k_proj.weight  ->  decoder.blocks.17.cross_attn.key.weight
decoder.blocks.17.cross_attn.key.weight 2 (1280, 1280)
model.decoder.layers.17.encoder_attn.v_proj.weight  ->  decoder.blocks.17.cross_attn.value.weight
decoder.blocks.17.cross_attn.value.weight 2 (1280, 1280)
model.decoder.layers.17.encoder_attn.v_proj.bias  ->  decoder.blocks.17.cross_attn.value.bias
decoder.blocks.17.cross_attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.17.encoder_attn.q_proj.weight  ->  decoder.blocks.17.cross_attn.query.weight
decoder.blocks.17.cross_attn.query.weight 2 (1280, 1280)
model.decoder.layers.17.encoder_attn.q_proj.bias  ->  decoder.blocks.17.cross_attn.query.bias
decoder.blocks.17.cross_attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.17.encoder_attn.out_proj.weight  ->  decoder.blocks.17.cross_attn.out.weight
decoder.blocks.17.cross_attn.out.weight 2 (1280, 1280)
model.decoder.layers.17.encoder_attn.out_proj.bias  ->  decoder.blocks.17.cross_attn.out.bias
decoder.blocks.17.cross_attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.17.encoder_attn_layer_norm.weight  ->  decoder.blocks.17.cross_attn_ln.weight
decoder.blocks.17.cross_attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.17.encoder_attn_layer_norm.bias  ->  decoder.blocks.17.cross_attn_ln.bias
decoder.blocks.17.cross_attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.17.fc1.weight  ->  decoder.blocks.17.mlp.0.weight
decoder.blocks.17.mlp.0.weight 2 (5120, 1280)
model.decoder.layers.17.fc1.bias  ->  decoder.blocks.17.mlp.0.bias
decoder.blocks.17.mlp.0.bias 1 (5120,)
  Converting to float32
model.decoder.layers.17.fc2.weight  ->  decoder.blocks.17.mlp.2.weight
decoder.blocks.17.mlp.2.weight 2 (1280, 5120)
model.decoder.layers.17.fc2.bias  ->  decoder.blocks.17.mlp.2.bias
decoder.blocks.17.mlp.2.bias 1 (1280,)
  Converting to float32
model.decoder.layers.17.final_layer_norm.weight  ->  decoder.blocks.17.mlp_ln.weight
decoder.blocks.17.mlp_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.17.final_layer_norm.bias  ->  decoder.blocks.17.mlp_ln.bias
decoder.blocks.17.mlp_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.18.self_attn.k_proj.weight  ->  decoder.blocks.18.attn.key.weight
decoder.blocks.18.attn.key.weight 2 (1280, 1280)
model.decoder.layers.18.self_attn.v_proj.weight  ->  decoder.blocks.18.attn.value.weight
decoder.blocks.18.attn.value.weight 2 (1280, 1280)
model.decoder.layers.18.self_attn.v_proj.bias  ->  decoder.blocks.18.attn.value.bias
decoder.blocks.18.attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.18.self_attn.q_proj.weight  ->  decoder.blocks.18.attn.query.weight
decoder.blocks.18.attn.query.weight 2 (1280, 1280)
model.decoder.layers.18.self_attn.q_proj.bias  ->  decoder.blocks.18.attn.query.bias
decoder.blocks.18.attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.18.self_attn.out_proj.weight  ->  decoder.blocks.18.attn.out.weight
decoder.blocks.18.attn.out.weight 2 (1280, 1280)
model.decoder.layers.18.self_attn.out_proj.bias  ->  decoder.blocks.18.attn.out.bias
decoder.blocks.18.attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.18.self_attn_layer_norm.weight  ->  decoder.blocks.18.attn_ln.weight
decoder.blocks.18.attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.18.self_attn_layer_norm.bias  ->  decoder.blocks.18.attn_ln.bias
decoder.blocks.18.attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.18.encoder_attn.k_proj.weight  ->  decoder.blocks.18.cross_attn.key.weight
decoder.blocks.18.cross_attn.key.weight 2 (1280, 1280)
model.decoder.layers.18.encoder_attn.v_proj.weight  ->  decoder.blocks.18.cross_attn.value.weight
decoder.blocks.18.cross_attn.value.weight 2 (1280, 1280)
model.decoder.layers.18.encoder_attn.v_proj.bias  ->  decoder.blocks.18.cross_attn.value.bias
decoder.blocks.18.cross_attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.18.encoder_attn.q_proj.weight  ->  decoder.blocks.18.cross_attn.query.weight
decoder.blocks.18.cross_attn.query.weight 2 (1280, 1280)
model.decoder.layers.18.encoder_attn.q_proj.bias  ->  decoder.blocks.18.cross_attn.query.bias
decoder.blocks.18.cross_attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.18.encoder_attn.out_proj.weight  ->  decoder.blocks.18.cross_attn.out.weight
decoder.blocks.18.cross_attn.out.weight 2 (1280, 1280)
model.decoder.layers.18.encoder_attn.out_proj.bias  ->  decoder.blocks.18.cross_attn.out.bias
decoder.blocks.18.cross_attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.18.encoder_attn_layer_norm.weight  ->  decoder.blocks.18.cross_attn_ln.weight
decoder.blocks.18.cross_attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.18.encoder_attn_layer_norm.bias  ->  decoder.blocks.18.cross_attn_ln.bias
decoder.blocks.18.cross_attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.18.fc1.weight  ->  decoder.blocks.18.mlp.0.weight
decoder.blocks.18.mlp.0.weight 2 (5120, 1280)
model.decoder.layers.18.fc1.bias  ->  decoder.blocks.18.mlp.0.bias
decoder.blocks.18.mlp.0.bias 1 (5120,)
  Converting to float32
model.decoder.layers.18.fc2.weight  ->  decoder.blocks.18.mlp.2.weight
decoder.blocks.18.mlp.2.weight 2 (1280, 5120)
model.decoder.layers.18.fc2.bias  ->  decoder.blocks.18.mlp.2.bias
decoder.blocks.18.mlp.2.bias 1 (1280,)
  Converting to float32
model.decoder.layers.18.final_layer_norm.weight  ->  decoder.blocks.18.mlp_ln.weight
decoder.blocks.18.mlp_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.18.final_layer_norm.bias  ->  decoder.blocks.18.mlp_ln.bias
decoder.blocks.18.mlp_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.19.self_attn.k_proj.weight  ->  decoder.blocks.19.attn.key.weight
decoder.blocks.19.attn.key.weight 2 (1280, 1280)
model.decoder.layers.19.self_attn.v_proj.weight  ->  decoder.blocks.19.attn.value.weight
decoder.blocks.19.attn.value.weight 2 (1280, 1280)
model.decoder.layers.19.self_attn.v_proj.bias  ->  decoder.blocks.19.attn.value.bias
decoder.blocks.19.attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.19.self_attn.q_proj.weight  ->  decoder.blocks.19.attn.query.weight
decoder.blocks.19.attn.query.weight 2 (1280, 1280)
model.decoder.layers.19.self_attn.q_proj.bias  ->  decoder.blocks.19.attn.query.bias
decoder.blocks.19.attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.19.self_attn.out_proj.weight  ->  decoder.blocks.19.attn.out.weight
decoder.blocks.19.attn.out.weight 2 (1280, 1280)
model.decoder.layers.19.self_attn.out_proj.bias  ->  decoder.blocks.19.attn.out.bias
decoder.blocks.19.attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.19.self_attn_layer_norm.weight  ->  decoder.blocks.19.attn_ln.weight
decoder.blocks.19.attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.19.self_attn_layer_norm.bias  ->  decoder.blocks.19.attn_ln.bias
decoder.blocks.19.attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.19.encoder_attn.k_proj.weight  ->  decoder.blocks.19.cross_attn.key.weight
decoder.blocks.19.cross_attn.key.weight 2 (1280, 1280)
model.decoder.layers.19.encoder_attn.v_proj.weight  ->  decoder.blocks.19.cross_attn.value.weight
decoder.blocks.19.cross_attn.value.weight 2 (1280, 1280)
model.decoder.layers.19.encoder_attn.v_proj.bias  ->  decoder.blocks.19.cross_attn.value.bias
decoder.blocks.19.cross_attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.19.encoder_attn.q_proj.weight  ->  decoder.blocks.19.cross_attn.query.weight
decoder.blocks.19.cross_attn.query.weight 2 (1280, 1280)
model.decoder.layers.19.encoder_attn.q_proj.bias  ->  decoder.blocks.19.cross_attn.query.bias
decoder.blocks.19.cross_attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.19.encoder_attn.out_proj.weight  ->  decoder.blocks.19.cross_attn.out.weight
decoder.blocks.19.cross_attn.out.weight 2 (1280, 1280)
model.decoder.layers.19.encoder_attn.out_proj.bias  ->  decoder.blocks.19.cross_attn.out.bias
decoder.blocks.19.cross_attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.19.encoder_attn_layer_norm.weight  ->  decoder.blocks.19.cross_attn_ln.weight
decoder.blocks.19.cross_attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.19.encoder_attn_layer_norm.bias  ->  decoder.blocks.19.cross_attn_ln.bias
decoder.blocks.19.cross_attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.19.fc1.weight  ->  decoder.blocks.19.mlp.0.weight
decoder.blocks.19.mlp.0.weight 2 (5120, 1280)
model.decoder.layers.19.fc1.bias  ->  decoder.blocks.19.mlp.0.bias
decoder.blocks.19.mlp.0.bias 1 (5120,)
  Converting to float32
model.decoder.layers.19.fc2.weight  ->  decoder.blocks.19.mlp.2.weight
decoder.blocks.19.mlp.2.weight 2 (1280, 5120)
model.decoder.layers.19.fc2.bias  ->  decoder.blocks.19.mlp.2.bias
decoder.blocks.19.mlp.2.bias 1 (1280,)
  Converting to float32
model.decoder.layers.19.final_layer_norm.weight  ->  decoder.blocks.19.mlp_ln.weight
decoder.blocks.19.mlp_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.19.final_layer_norm.bias  ->  decoder.blocks.19.mlp_ln.bias
decoder.blocks.19.mlp_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.20.self_attn.k_proj.weight  ->  decoder.blocks.20.attn.key.weight
decoder.blocks.20.attn.key.weight 2 (1280, 1280)
model.decoder.layers.20.self_attn.v_proj.weight  ->  decoder.blocks.20.attn.value.weight
decoder.blocks.20.attn.value.weight 2 (1280, 1280)
model.decoder.layers.20.self_attn.v_proj.bias  ->  decoder.blocks.20.attn.value.bias
decoder.blocks.20.attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.20.self_attn.q_proj.weight  ->  decoder.blocks.20.attn.query.weight
decoder.blocks.20.attn.query.weight 2 (1280, 1280)
model.decoder.layers.20.self_attn.q_proj.bias  ->  decoder.blocks.20.attn.query.bias
decoder.blocks.20.attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.20.self_attn.out_proj.weight  ->  decoder.blocks.20.attn.out.weight
decoder.blocks.20.attn.out.weight 2 (1280, 1280)
model.decoder.layers.20.self_attn.out_proj.bias  ->  decoder.blocks.20.attn.out.bias
decoder.blocks.20.attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.20.self_attn_layer_norm.weight  ->  decoder.blocks.20.attn_ln.weight
decoder.blocks.20.attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.20.self_attn_layer_norm.bias  ->  decoder.blocks.20.attn_ln.bias
decoder.blocks.20.attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.20.encoder_attn.k_proj.weight  ->  decoder.blocks.20.cross_attn.key.weight
decoder.blocks.20.cross_attn.key.weight 2 (1280, 1280)
model.decoder.layers.20.encoder_attn.v_proj.weight  ->  decoder.blocks.20.cross_attn.value.weight
decoder.blocks.20.cross_attn.value.weight 2 (1280, 1280)
model.decoder.layers.20.encoder_attn.v_proj.bias  ->  decoder.blocks.20.cross_attn.value.bias
decoder.blocks.20.cross_attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.20.encoder_attn.q_proj.weight  ->  decoder.blocks.20.cross_attn.query.weight
decoder.blocks.20.cross_attn.query.weight 2 (1280, 1280)
model.decoder.layers.20.encoder_attn.q_proj.bias  ->  decoder.blocks.20.cross_attn.query.bias
decoder.blocks.20.cross_attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.20.encoder_attn.out_proj.weight  ->  decoder.blocks.20.cross_attn.out.weight
decoder.blocks.20.cross_attn.out.weight 2 (1280, 1280)
model.decoder.layers.20.encoder_attn.out_proj.bias  ->  decoder.blocks.20.cross_attn.out.bias
decoder.blocks.20.cross_attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.20.encoder_attn_layer_norm.weight  ->  decoder.blocks.20.cross_attn_ln.weight
decoder.blocks.20.cross_attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.20.encoder_attn_layer_norm.bias  ->  decoder.blocks.20.cross_attn_ln.bias
decoder.blocks.20.cross_attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.20.fc1.weight  ->  decoder.blocks.20.mlp.0.weight
decoder.blocks.20.mlp.0.weight 2 (5120, 1280)
model.decoder.layers.20.fc1.bias  ->  decoder.blocks.20.mlp.0.bias
decoder.blocks.20.mlp.0.bias 1 (5120,)
  Converting to float32
model.decoder.layers.20.fc2.weight  ->  decoder.blocks.20.mlp.2.weight
decoder.blocks.20.mlp.2.weight 2 (1280, 5120)
model.decoder.layers.20.fc2.bias  ->  decoder.blocks.20.mlp.2.bias
decoder.blocks.20.mlp.2.bias 1 (1280,)
  Converting to float32
model.decoder.layers.20.final_layer_norm.weight  ->  decoder.blocks.20.mlp_ln.weight
decoder.blocks.20.mlp_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.20.final_layer_norm.bias  ->  decoder.blocks.20.mlp_ln.bias
decoder.blocks.20.mlp_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.21.self_attn.k_proj.weight  ->  decoder.blocks.21.attn.key.weight
decoder.blocks.21.attn.key.weight 2 (1280, 1280)
model.decoder.layers.21.self_attn.v_proj.weight  ->  decoder.blocks.21.attn.value.weight
decoder.blocks.21.attn.value.weight 2 (1280, 1280)
model.decoder.layers.21.self_attn.v_proj.bias  ->  decoder.blocks.21.attn.value.bias
decoder.blocks.21.attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.21.self_attn.q_proj.weight  ->  decoder.blocks.21.attn.query.weight
decoder.blocks.21.attn.query.weight 2 (1280, 1280)
model.decoder.layers.21.self_attn.q_proj.bias  ->  decoder.blocks.21.attn.query.bias
decoder.blocks.21.attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.21.self_attn.out_proj.weight  ->  decoder.blocks.21.attn.out.weight
decoder.blocks.21.attn.out.weight 2 (1280, 1280)
model.decoder.layers.21.self_attn.out_proj.bias  ->  decoder.blocks.21.attn.out.bias
decoder.blocks.21.attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.21.self_attn_layer_norm.weight  ->  decoder.blocks.21.attn_ln.weight
decoder.blocks.21.attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.21.self_attn_layer_norm.bias  ->  decoder.blocks.21.attn_ln.bias
decoder.blocks.21.attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.21.encoder_attn.k_proj.weight  ->  decoder.blocks.21.cross_attn.key.weight
decoder.blocks.21.cross_attn.key.weight 2 (1280, 1280)
model.decoder.layers.21.encoder_attn.v_proj.weight  ->  decoder.blocks.21.cross_attn.value.weight
decoder.blocks.21.cross_attn.value.weight 2 (1280, 1280)
model.decoder.layers.21.encoder_attn.v_proj.bias  ->  decoder.blocks.21.cross_attn.value.bias
decoder.blocks.21.cross_attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.21.encoder_attn.q_proj.weight  ->  decoder.blocks.21.cross_attn.query.weight
decoder.blocks.21.cross_attn.query.weight 2 (1280, 1280)
model.decoder.layers.21.encoder_attn.q_proj.bias  ->  decoder.blocks.21.cross_attn.query.bias
decoder.blocks.21.cross_attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.21.encoder_attn.out_proj.weight  ->  decoder.blocks.21.cross_attn.out.weight
decoder.blocks.21.cross_attn.out.weight 2 (1280, 1280)
model.decoder.layers.21.encoder_attn.out_proj.bias  ->  decoder.blocks.21.cross_attn.out.bias
decoder.blocks.21.cross_attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.21.encoder_attn_layer_norm.weight  ->  decoder.blocks.21.cross_attn_ln.weight
decoder.blocks.21.cross_attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.21.encoder_attn_layer_norm.bias  ->  decoder.blocks.21.cross_attn_ln.bias
decoder.blocks.21.cross_attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.21.fc1.weight  ->  decoder.blocks.21.mlp.0.weight
decoder.blocks.21.mlp.0.weight 2 (5120, 1280)
model.decoder.layers.21.fc1.bias  ->  decoder.blocks.21.mlp.0.bias
decoder.blocks.21.mlp.0.bias 1 (5120,)
  Converting to float32
model.decoder.layers.21.fc2.weight  ->  decoder.blocks.21.mlp.2.weight
decoder.blocks.21.mlp.2.weight 2 (1280, 5120)
model.decoder.layers.21.fc2.bias  ->  decoder.blocks.21.mlp.2.bias
decoder.blocks.21.mlp.2.bias 1 (1280,)
  Converting to float32
model.decoder.layers.21.final_layer_norm.weight  ->  decoder.blocks.21.mlp_ln.weight
decoder.blocks.21.mlp_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.21.final_layer_norm.bias  ->  decoder.blocks.21.mlp_ln.bias
decoder.blocks.21.mlp_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.22.self_attn.k_proj.weight  ->  decoder.blocks.22.attn.key.weight
decoder.blocks.22.attn.key.weight 2 (1280, 1280)
model.decoder.layers.22.self_attn.v_proj.weight  ->  decoder.blocks.22.attn.value.weight
decoder.blocks.22.attn.value.weight 2 (1280, 1280)
model.decoder.layers.22.self_attn.v_proj.bias  ->  decoder.blocks.22.attn.value.bias
decoder.blocks.22.attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.22.self_attn.q_proj.weight  ->  decoder.blocks.22.attn.query.weight
decoder.blocks.22.attn.query.weight 2 (1280, 1280)
model.decoder.layers.22.self_attn.q_proj.bias  ->  decoder.blocks.22.attn.query.bias
decoder.blocks.22.attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.22.self_attn.out_proj.weight  ->  decoder.blocks.22.attn.out.weight
decoder.blocks.22.attn.out.weight 2 (1280, 1280)
model.decoder.layers.22.self_attn.out_proj.bias  ->  decoder.blocks.22.attn.out.bias
decoder.blocks.22.attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.22.self_attn_layer_norm.weight  ->  decoder.blocks.22.attn_ln.weight
decoder.blocks.22.attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.22.self_attn_layer_norm.bias  ->  decoder.blocks.22.attn_ln.bias
decoder.blocks.22.attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.22.encoder_attn.k_proj.weight  ->  decoder.blocks.22.cross_attn.key.weight
decoder.blocks.22.cross_attn.key.weight 2 (1280, 1280)
model.decoder.layers.22.encoder_attn.v_proj.weight  ->  decoder.blocks.22.cross_attn.value.weight
decoder.blocks.22.cross_attn.value.weight 2 (1280, 1280)
model.decoder.layers.22.encoder_attn.v_proj.bias  ->  decoder.blocks.22.cross_attn.value.bias
decoder.blocks.22.cross_attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.22.encoder_attn.q_proj.weight  ->  decoder.blocks.22.cross_attn.query.weight
decoder.blocks.22.cross_attn.query.weight 2 (1280, 1280)
model.decoder.layers.22.encoder_attn.q_proj.bias  ->  decoder.blocks.22.cross_attn.query.bias
decoder.blocks.22.cross_attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.22.encoder_attn.out_proj.weight  ->  decoder.blocks.22.cross_attn.out.weight
decoder.blocks.22.cross_attn.out.weight 2 (1280, 1280)
model.decoder.layers.22.encoder_attn.out_proj.bias  ->  decoder.blocks.22.cross_attn.out.bias
decoder.blocks.22.cross_attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.22.encoder_attn_layer_norm.weight  ->  decoder.blocks.22.cross_attn_ln.weight
decoder.blocks.22.cross_attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.22.encoder_attn_layer_norm.bias  ->  decoder.blocks.22.cross_attn_ln.bias
decoder.blocks.22.cross_attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.22.fc1.weight  ->  decoder.blocks.22.mlp.0.weight
decoder.blocks.22.mlp.0.weight 2 (5120, 1280)
model.decoder.layers.22.fc1.bias  ->  decoder.blocks.22.mlp.0.bias
decoder.blocks.22.mlp.0.bias 1 (5120,)
  Converting to float32
model.decoder.layers.22.fc2.weight  ->  decoder.blocks.22.mlp.2.weight
decoder.blocks.22.mlp.2.weight 2 (1280, 5120)
model.decoder.layers.22.fc2.bias  ->  decoder.blocks.22.mlp.2.bias
decoder.blocks.22.mlp.2.bias 1 (1280,)
  Converting to float32
model.decoder.layers.22.final_layer_norm.weight  ->  decoder.blocks.22.mlp_ln.weight
decoder.blocks.22.mlp_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.22.final_layer_norm.bias  ->  decoder.blocks.22.mlp_ln.bias
decoder.blocks.22.mlp_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.23.self_attn.k_proj.weight  ->  decoder.blocks.23.attn.key.weight
decoder.blocks.23.attn.key.weight 2 (1280, 1280)
model.decoder.layers.23.self_attn.v_proj.weight  ->  decoder.blocks.23.attn.value.weight
decoder.blocks.23.attn.value.weight 2 (1280, 1280)
model.decoder.layers.23.self_attn.v_proj.bias  ->  decoder.blocks.23.attn.value.bias
decoder.blocks.23.attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.23.self_attn.q_proj.weight  ->  decoder.blocks.23.attn.query.weight
decoder.blocks.23.attn.query.weight 2 (1280, 1280)
model.decoder.layers.23.self_attn.q_proj.bias  ->  decoder.blocks.23.attn.query.bias
decoder.blocks.23.attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.23.self_attn.out_proj.weight  ->  decoder.blocks.23.attn.out.weight
decoder.blocks.23.attn.out.weight 2 (1280, 1280)
model.decoder.layers.23.self_attn.out_proj.bias  ->  decoder.blocks.23.attn.out.bias
decoder.blocks.23.attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.23.self_attn_layer_norm.weight  ->  decoder.blocks.23.attn_ln.weight
decoder.blocks.23.attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.23.self_attn_layer_norm.bias  ->  decoder.blocks.23.attn_ln.bias
decoder.blocks.23.attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.23.encoder_attn.k_proj.weight  ->  decoder.blocks.23.cross_attn.key.weight
decoder.blocks.23.cross_attn.key.weight 2 (1280, 1280)
model.decoder.layers.23.encoder_attn.v_proj.weight  ->  decoder.blocks.23.cross_attn.value.weight
decoder.blocks.23.cross_attn.value.weight 2 (1280, 1280)
model.decoder.layers.23.encoder_attn.v_proj.bias  ->  decoder.blocks.23.cross_attn.value.bias
decoder.blocks.23.cross_attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.23.encoder_attn.q_proj.weight  ->  decoder.blocks.23.cross_attn.query.weight
decoder.blocks.23.cross_attn.query.weight 2 (1280, 1280)
model.decoder.layers.23.encoder_attn.q_proj.bias  ->  decoder.blocks.23.cross_attn.query.bias
decoder.blocks.23.cross_attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.23.encoder_attn.out_proj.weight  ->  decoder.blocks.23.cross_attn.out.weight
decoder.blocks.23.cross_attn.out.weight 2 (1280, 1280)
model.decoder.layers.23.encoder_attn.out_proj.bias  ->  decoder.blocks.23.cross_attn.out.bias
decoder.blocks.23.cross_attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.23.encoder_attn_layer_norm.weight  ->  decoder.blocks.23.cross_attn_ln.weight
decoder.blocks.23.cross_attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.23.encoder_attn_layer_norm.bias  ->  decoder.blocks.23.cross_attn_ln.bias
decoder.blocks.23.cross_attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.23.fc1.weight  ->  decoder.blocks.23.mlp.0.weight
decoder.blocks.23.mlp.0.weight 2 (5120, 1280)
model.decoder.layers.23.fc1.bias  ->  decoder.blocks.23.mlp.0.bias
decoder.blocks.23.mlp.0.bias 1 (5120,)
  Converting to float32
model.decoder.layers.23.fc2.weight  ->  decoder.blocks.23.mlp.2.weight
decoder.blocks.23.mlp.2.weight 2 (1280, 5120)
model.decoder.layers.23.fc2.bias  ->  decoder.blocks.23.mlp.2.bias
decoder.blocks.23.mlp.2.bias 1 (1280,)
  Converting to float32
model.decoder.layers.23.final_layer_norm.weight  ->  decoder.blocks.23.mlp_ln.weight
decoder.blocks.23.mlp_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.23.final_layer_norm.bias  ->  decoder.blocks.23.mlp_ln.bias
decoder.blocks.23.mlp_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.24.self_attn.k_proj.weight  ->  decoder.blocks.24.attn.key.weight
decoder.blocks.24.attn.key.weight 2 (1280, 1280)
model.decoder.layers.24.self_attn.v_proj.weight  ->  decoder.blocks.24.attn.value.weight
decoder.blocks.24.attn.value.weight 2 (1280, 1280)
model.decoder.layers.24.self_attn.v_proj.bias  ->  decoder.blocks.24.attn.value.bias
decoder.blocks.24.attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.24.self_attn.q_proj.weight  ->  decoder.blocks.24.attn.query.weight
decoder.blocks.24.attn.query.weight 2 (1280, 1280)
model.decoder.layers.24.self_attn.q_proj.bias  ->  decoder.blocks.24.attn.query.bias
decoder.blocks.24.attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.24.self_attn.out_proj.weight  ->  decoder.blocks.24.attn.out.weight
decoder.blocks.24.attn.out.weight 2 (1280, 1280)
model.decoder.layers.24.self_attn.out_proj.bias  ->  decoder.blocks.24.attn.out.bias
decoder.blocks.24.attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.24.self_attn_layer_norm.weight  ->  decoder.blocks.24.attn_ln.weight
decoder.blocks.24.attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.24.self_attn_layer_norm.bias  ->  decoder.blocks.24.attn_ln.bias
decoder.blocks.24.attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.24.encoder_attn.k_proj.weight  ->  decoder.blocks.24.cross_attn.key.weight
decoder.blocks.24.cross_attn.key.weight 2 (1280, 1280)
model.decoder.layers.24.encoder_attn.v_proj.weight  ->  decoder.blocks.24.cross_attn.value.weight
decoder.blocks.24.cross_attn.value.weight 2 (1280, 1280)
model.decoder.layers.24.encoder_attn.v_proj.bias  ->  decoder.blocks.24.cross_attn.value.bias
decoder.blocks.24.cross_attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.24.encoder_attn.q_proj.weight  ->  decoder.blocks.24.cross_attn.query.weight
decoder.blocks.24.cross_attn.query.weight 2 (1280, 1280)
model.decoder.layers.24.encoder_attn.q_proj.bias  ->  decoder.blocks.24.cross_attn.query.bias
decoder.blocks.24.cross_attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.24.encoder_attn.out_proj.weight  ->  decoder.blocks.24.cross_attn.out.weight
decoder.blocks.24.cross_attn.out.weight 2 (1280, 1280)
model.decoder.layers.24.encoder_attn.out_proj.bias  ->  decoder.blocks.24.cross_attn.out.bias
decoder.blocks.24.cross_attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.24.encoder_attn_layer_norm.weight  ->  decoder.blocks.24.cross_attn_ln.weight
decoder.blocks.24.cross_attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.24.encoder_attn_layer_norm.bias  ->  decoder.blocks.24.cross_attn_ln.bias
decoder.blocks.24.cross_attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.24.fc1.weight  ->  decoder.blocks.24.mlp.0.weight
decoder.blocks.24.mlp.0.weight 2 (5120, 1280)
model.decoder.layers.24.fc1.bias  ->  decoder.blocks.24.mlp.0.bias
decoder.blocks.24.mlp.0.bias 1 (5120,)
  Converting to float32
model.decoder.layers.24.fc2.weight  ->  decoder.blocks.24.mlp.2.weight
decoder.blocks.24.mlp.2.weight 2 (1280, 5120)
model.decoder.layers.24.fc2.bias  ->  decoder.blocks.24.mlp.2.bias
decoder.blocks.24.mlp.2.bias 1 (1280,)
  Converting to float32
model.decoder.layers.24.final_layer_norm.weight  ->  decoder.blocks.24.mlp_ln.weight
decoder.blocks.24.mlp_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.24.final_layer_norm.bias  ->  decoder.blocks.24.mlp_ln.bias
decoder.blocks.24.mlp_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.25.self_attn.k_proj.weight  ->  decoder.blocks.25.attn.key.weight
decoder.blocks.25.attn.key.weight 2 (1280, 1280)
model.decoder.layers.25.self_attn.v_proj.weight  ->  decoder.blocks.25.attn.value.weight
decoder.blocks.25.attn.value.weight 2 (1280, 1280)
model.decoder.layers.25.self_attn.v_proj.bias  ->  decoder.blocks.25.attn.value.bias
decoder.blocks.25.attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.25.self_attn.q_proj.weight  ->  decoder.blocks.25.attn.query.weight
decoder.blocks.25.attn.query.weight 2 (1280, 1280)
model.decoder.layers.25.self_attn.q_proj.bias  ->  decoder.blocks.25.attn.query.bias
decoder.blocks.25.attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.25.self_attn.out_proj.weight  ->  decoder.blocks.25.attn.out.weight
decoder.blocks.25.attn.out.weight 2 (1280, 1280)
model.decoder.layers.25.self_attn.out_proj.bias  ->  decoder.blocks.25.attn.out.bias
decoder.blocks.25.attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.25.self_attn_layer_norm.weight  ->  decoder.blocks.25.attn_ln.weight
decoder.blocks.25.attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.25.self_attn_layer_norm.bias  ->  decoder.blocks.25.attn_ln.bias
decoder.blocks.25.attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.25.encoder_attn.k_proj.weight  ->  decoder.blocks.25.cross_attn.key.weight
decoder.blocks.25.cross_attn.key.weight 2 (1280, 1280)
model.decoder.layers.25.encoder_attn.v_proj.weight  ->  decoder.blocks.25.cross_attn.value.weight
decoder.blocks.25.cross_attn.value.weight 2 (1280, 1280)
model.decoder.layers.25.encoder_attn.v_proj.bias  ->  decoder.blocks.25.cross_attn.value.bias
decoder.blocks.25.cross_attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.25.encoder_attn.q_proj.weight  ->  decoder.blocks.25.cross_attn.query.weight
decoder.blocks.25.cross_attn.query.weight 2 (1280, 1280)
model.decoder.layers.25.encoder_attn.q_proj.bias  ->  decoder.blocks.25.cross_attn.query.bias
decoder.blocks.25.cross_attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.25.encoder_attn.out_proj.weight  ->  decoder.blocks.25.cross_attn.out.weight
decoder.blocks.25.cross_attn.out.weight 2 (1280, 1280)
model.decoder.layers.25.encoder_attn.out_proj.bias  ->  decoder.blocks.25.cross_attn.out.bias
decoder.blocks.25.cross_attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.25.encoder_attn_layer_norm.weight  ->  decoder.blocks.25.cross_attn_ln.weight
decoder.blocks.25.cross_attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.25.encoder_attn_layer_norm.bias  ->  decoder.blocks.25.cross_attn_ln.bias
decoder.blocks.25.cross_attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.25.fc1.weight  ->  decoder.blocks.25.mlp.0.weight
decoder.blocks.25.mlp.0.weight 2 (5120, 1280)
model.decoder.layers.25.fc1.bias  ->  decoder.blocks.25.mlp.0.bias
decoder.blocks.25.mlp.0.bias 1 (5120,)
  Converting to float32
model.decoder.layers.25.fc2.weight  ->  decoder.blocks.25.mlp.2.weight
decoder.blocks.25.mlp.2.weight 2 (1280, 5120)
model.decoder.layers.25.fc2.bias  ->  decoder.blocks.25.mlp.2.bias
decoder.blocks.25.mlp.2.bias 1 (1280,)
  Converting to float32
model.decoder.layers.25.final_layer_norm.weight  ->  decoder.blocks.25.mlp_ln.weight
decoder.blocks.25.mlp_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.25.final_layer_norm.bias  ->  decoder.blocks.25.mlp_ln.bias
decoder.blocks.25.mlp_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.26.self_attn.k_proj.weight  ->  decoder.blocks.26.attn.key.weight
decoder.blocks.26.attn.key.weight 2 (1280, 1280)
model.decoder.layers.26.self_attn.v_proj.weight  ->  decoder.blocks.26.attn.value.weight
decoder.blocks.26.attn.value.weight 2 (1280, 1280)
model.decoder.layers.26.self_attn.v_proj.bias  ->  decoder.blocks.26.attn.value.bias
decoder.blocks.26.attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.26.self_attn.q_proj.weight  ->  decoder.blocks.26.attn.query.weight
decoder.blocks.26.attn.query.weight 2 (1280, 1280)
model.decoder.layers.26.self_attn.q_proj.bias  ->  decoder.blocks.26.attn.query.bias
decoder.blocks.26.attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.26.self_attn.out_proj.weight  ->  decoder.blocks.26.attn.out.weight
decoder.blocks.26.attn.out.weight 2 (1280, 1280)
model.decoder.layers.26.self_attn.out_proj.bias  ->  decoder.blocks.26.attn.out.bias
decoder.blocks.26.attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.26.self_attn_layer_norm.weight  ->  decoder.blocks.26.attn_ln.weight
decoder.blocks.26.attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.26.self_attn_layer_norm.bias  ->  decoder.blocks.26.attn_ln.bias
decoder.blocks.26.attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.26.encoder_attn.k_proj.weight  ->  decoder.blocks.26.cross_attn.key.weight
decoder.blocks.26.cross_attn.key.weight 2 (1280, 1280)
model.decoder.layers.26.encoder_attn.v_proj.weight  ->  decoder.blocks.26.cross_attn.value.weight
decoder.blocks.26.cross_attn.value.weight 2 (1280, 1280)
model.decoder.layers.26.encoder_attn.v_proj.bias  ->  decoder.blocks.26.cross_attn.value.bias
decoder.blocks.26.cross_attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.26.encoder_attn.q_proj.weight  ->  decoder.blocks.26.cross_attn.query.weight
decoder.blocks.26.cross_attn.query.weight 2 (1280, 1280)
model.decoder.layers.26.encoder_attn.q_proj.bias  ->  decoder.blocks.26.cross_attn.query.bias
decoder.blocks.26.cross_attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.26.encoder_attn.out_proj.weight  ->  decoder.blocks.26.cross_attn.out.weight
decoder.blocks.26.cross_attn.out.weight 2 (1280, 1280)
model.decoder.layers.26.encoder_attn.out_proj.bias  ->  decoder.blocks.26.cross_attn.out.bias
decoder.blocks.26.cross_attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.26.encoder_attn_layer_norm.weight  ->  decoder.blocks.26.cross_attn_ln.weight
decoder.blocks.26.cross_attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.26.encoder_attn_layer_norm.bias  ->  decoder.blocks.26.cross_attn_ln.bias
decoder.blocks.26.cross_attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.26.fc1.weight  ->  decoder.blocks.26.mlp.0.weight
decoder.blocks.26.mlp.0.weight 2 (5120, 1280)
model.decoder.layers.26.fc1.bias  ->  decoder.blocks.26.mlp.0.bias
decoder.blocks.26.mlp.0.bias 1 (5120,)
  Converting to float32
model.decoder.layers.26.fc2.weight  ->  decoder.blocks.26.mlp.2.weight
decoder.blocks.26.mlp.2.weight 2 (1280, 5120)
model.decoder.layers.26.fc2.bias  ->  decoder.blocks.26.mlp.2.bias
decoder.blocks.26.mlp.2.bias 1 (1280,)
  Converting to float32
model.decoder.layers.26.final_layer_norm.weight  ->  decoder.blocks.26.mlp_ln.weight
decoder.blocks.26.mlp_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.26.final_layer_norm.bias  ->  decoder.blocks.26.mlp_ln.bias
decoder.blocks.26.mlp_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.27.self_attn.k_proj.weight  ->  decoder.blocks.27.attn.key.weight
decoder.blocks.27.attn.key.weight 2 (1280, 1280)
model.decoder.layers.27.self_attn.v_proj.weight  ->  decoder.blocks.27.attn.value.weight
decoder.blocks.27.attn.value.weight 2 (1280, 1280)
model.decoder.layers.27.self_attn.v_proj.bias  ->  decoder.blocks.27.attn.value.bias
decoder.blocks.27.attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.27.self_attn.q_proj.weight  ->  decoder.blocks.27.attn.query.weight
decoder.blocks.27.attn.query.weight 2 (1280, 1280)
model.decoder.layers.27.self_attn.q_proj.bias  ->  decoder.blocks.27.attn.query.bias
decoder.blocks.27.attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.27.self_attn.out_proj.weight  ->  decoder.blocks.27.attn.out.weight
decoder.blocks.27.attn.out.weight 2 (1280, 1280)
model.decoder.layers.27.self_attn.out_proj.bias  ->  decoder.blocks.27.attn.out.bias
decoder.blocks.27.attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.27.self_attn_layer_norm.weight  ->  decoder.blocks.27.attn_ln.weight
decoder.blocks.27.attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.27.self_attn_layer_norm.bias  ->  decoder.blocks.27.attn_ln.bias
decoder.blocks.27.attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.27.encoder_attn.k_proj.weight  ->  decoder.blocks.27.cross_attn.key.weight
decoder.blocks.27.cross_attn.key.weight 2 (1280, 1280)
model.decoder.layers.27.encoder_attn.v_proj.weight  ->  decoder.blocks.27.cross_attn.value.weight
decoder.blocks.27.cross_attn.value.weight 2 (1280, 1280)
model.decoder.layers.27.encoder_attn.v_proj.bias  ->  decoder.blocks.27.cross_attn.value.bias
decoder.blocks.27.cross_attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.27.encoder_attn.q_proj.weight  ->  decoder.blocks.27.cross_attn.query.weight
decoder.blocks.27.cross_attn.query.weight 2 (1280, 1280)
model.decoder.layers.27.encoder_attn.q_proj.bias  ->  decoder.blocks.27.cross_attn.query.bias
decoder.blocks.27.cross_attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.27.encoder_attn.out_proj.weight  ->  decoder.blocks.27.cross_attn.out.weight
decoder.blocks.27.cross_attn.out.weight 2 (1280, 1280)
model.decoder.layers.27.encoder_attn.out_proj.bias  ->  decoder.blocks.27.cross_attn.out.bias
decoder.blocks.27.cross_attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.27.encoder_attn_layer_norm.weight  ->  decoder.blocks.27.cross_attn_ln.weight
decoder.blocks.27.cross_attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.27.encoder_attn_layer_norm.bias  ->  decoder.blocks.27.cross_attn_ln.bias
decoder.blocks.27.cross_attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.27.fc1.weight  ->  decoder.blocks.27.mlp.0.weight
decoder.blocks.27.mlp.0.weight 2 (5120, 1280)
model.decoder.layers.27.fc1.bias  ->  decoder.blocks.27.mlp.0.bias
decoder.blocks.27.mlp.0.bias 1 (5120,)
  Converting to float32
model.decoder.layers.27.fc2.weight  ->  decoder.blocks.27.mlp.2.weight
decoder.blocks.27.mlp.2.weight 2 (1280, 5120)
model.decoder.layers.27.fc2.bias  ->  decoder.blocks.27.mlp.2.bias
decoder.blocks.27.mlp.2.bias 1 (1280,)
  Converting to float32
model.decoder.layers.27.final_layer_norm.weight  ->  decoder.blocks.27.mlp_ln.weight
decoder.blocks.27.mlp_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.27.final_layer_norm.bias  ->  decoder.blocks.27.mlp_ln.bias
decoder.blocks.27.mlp_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.28.self_attn.k_proj.weight  ->  decoder.blocks.28.attn.key.weight
decoder.blocks.28.attn.key.weight 2 (1280, 1280)
model.decoder.layers.28.self_attn.v_proj.weight  ->  decoder.blocks.28.attn.value.weight
decoder.blocks.28.attn.value.weight 2 (1280, 1280)
model.decoder.layers.28.self_attn.v_proj.bias  ->  decoder.blocks.28.attn.value.bias
decoder.blocks.28.attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.28.self_attn.q_proj.weight  ->  decoder.blocks.28.attn.query.weight
decoder.blocks.28.attn.query.weight 2 (1280, 1280)
model.decoder.layers.28.self_attn.q_proj.bias  ->  decoder.blocks.28.attn.query.bias
decoder.blocks.28.attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.28.self_attn.out_proj.weight  ->  decoder.blocks.28.attn.out.weight
decoder.blocks.28.attn.out.weight 2 (1280, 1280)
model.decoder.layers.28.self_attn.out_proj.bias  ->  decoder.blocks.28.attn.out.bias
decoder.blocks.28.attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.28.self_attn_layer_norm.weight  ->  decoder.blocks.28.attn_ln.weight
decoder.blocks.28.attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.28.self_attn_layer_norm.bias  ->  decoder.blocks.28.attn_ln.bias
decoder.blocks.28.attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.28.encoder_attn.k_proj.weight  ->  decoder.blocks.28.cross_attn.key.weight
decoder.blocks.28.cross_attn.key.weight 2 (1280, 1280)
model.decoder.layers.28.encoder_attn.v_proj.weight  ->  decoder.blocks.28.cross_attn.value.weight
decoder.blocks.28.cross_attn.value.weight 2 (1280, 1280)
model.decoder.layers.28.encoder_attn.v_proj.bias  ->  decoder.blocks.28.cross_attn.value.bias
decoder.blocks.28.cross_attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.28.encoder_attn.q_proj.weight  ->  decoder.blocks.28.cross_attn.query.weight
decoder.blocks.28.cross_attn.query.weight 2 (1280, 1280)
model.decoder.layers.28.encoder_attn.q_proj.bias  ->  decoder.blocks.28.cross_attn.query.bias
decoder.blocks.28.cross_attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.28.encoder_attn.out_proj.weight  ->  decoder.blocks.28.cross_attn.out.weight
decoder.blocks.28.cross_attn.out.weight 2 (1280, 1280)
model.decoder.layers.28.encoder_attn.out_proj.bias  ->  decoder.blocks.28.cross_attn.out.bias
decoder.blocks.28.cross_attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.28.encoder_attn_layer_norm.weight  ->  decoder.blocks.28.cross_attn_ln.weight
decoder.blocks.28.cross_attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.28.encoder_attn_layer_norm.bias  ->  decoder.blocks.28.cross_attn_ln.bias
decoder.blocks.28.cross_attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.28.fc1.weight  ->  decoder.blocks.28.mlp.0.weight
decoder.blocks.28.mlp.0.weight 2 (5120, 1280)
model.decoder.layers.28.fc1.bias  ->  decoder.blocks.28.mlp.0.bias
decoder.blocks.28.mlp.0.bias 1 (5120,)
  Converting to float32
model.decoder.layers.28.fc2.weight  ->  decoder.blocks.28.mlp.2.weight
decoder.blocks.28.mlp.2.weight 2 (1280, 5120)
model.decoder.layers.28.fc2.bias  ->  decoder.blocks.28.mlp.2.bias
decoder.blocks.28.mlp.2.bias 1 (1280,)
  Converting to float32
model.decoder.layers.28.final_layer_norm.weight  ->  decoder.blocks.28.mlp_ln.weight
decoder.blocks.28.mlp_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.28.final_layer_norm.bias  ->  decoder.blocks.28.mlp_ln.bias
decoder.blocks.28.mlp_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.29.self_attn.k_proj.weight  ->  decoder.blocks.29.attn.key.weight
decoder.blocks.29.attn.key.weight 2 (1280, 1280)
model.decoder.layers.29.self_attn.v_proj.weight  ->  decoder.blocks.29.attn.value.weight
decoder.blocks.29.attn.value.weight 2 (1280, 1280)
model.decoder.layers.29.self_attn.v_proj.bias  ->  decoder.blocks.29.attn.value.bias
decoder.blocks.29.attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.29.self_attn.q_proj.weight  ->  decoder.blocks.29.attn.query.weight
decoder.blocks.29.attn.query.weight 2 (1280, 1280)
model.decoder.layers.29.self_attn.q_proj.bias  ->  decoder.blocks.29.attn.query.bias
decoder.blocks.29.attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.29.self_attn.out_proj.weight  ->  decoder.blocks.29.attn.out.weight
decoder.blocks.29.attn.out.weight 2 (1280, 1280)
model.decoder.layers.29.self_attn.out_proj.bias  ->  decoder.blocks.29.attn.out.bias
decoder.blocks.29.attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.29.self_attn_layer_norm.weight  ->  decoder.blocks.29.attn_ln.weight
decoder.blocks.29.attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.29.self_attn_layer_norm.bias  ->  decoder.blocks.29.attn_ln.bias
decoder.blocks.29.attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.29.encoder_attn.k_proj.weight  ->  decoder.blocks.29.cross_attn.key.weight
decoder.blocks.29.cross_attn.key.weight 2 (1280, 1280)
model.decoder.layers.29.encoder_attn.v_proj.weight  ->  decoder.blocks.29.cross_attn.value.weight
decoder.blocks.29.cross_attn.value.weight 2 (1280, 1280)
model.decoder.layers.29.encoder_attn.v_proj.bias  ->  decoder.blocks.29.cross_attn.value.bias
decoder.blocks.29.cross_attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.29.encoder_attn.q_proj.weight  ->  decoder.blocks.29.cross_attn.query.weight
decoder.blocks.29.cross_attn.query.weight 2 (1280, 1280)
model.decoder.layers.29.encoder_attn.q_proj.bias  ->  decoder.blocks.29.cross_attn.query.bias
decoder.blocks.29.cross_attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.29.encoder_attn.out_proj.weight  ->  decoder.blocks.29.cross_attn.out.weight
decoder.blocks.29.cross_attn.out.weight 2 (1280, 1280)
model.decoder.layers.29.encoder_attn.out_proj.bias  ->  decoder.blocks.29.cross_attn.out.bias
decoder.blocks.29.cross_attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.29.encoder_attn_layer_norm.weight  ->  decoder.blocks.29.cross_attn_ln.weight
decoder.blocks.29.cross_attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.29.encoder_attn_layer_norm.bias  ->  decoder.blocks.29.cross_attn_ln.bias
decoder.blocks.29.cross_attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.29.fc1.weight  ->  decoder.blocks.29.mlp.0.weight
decoder.blocks.29.mlp.0.weight 2 (5120, 1280)
model.decoder.layers.29.fc1.bias  ->  decoder.blocks.29.mlp.0.bias
decoder.blocks.29.mlp.0.bias 1 (5120,)
  Converting to float32
model.decoder.layers.29.fc2.weight  ->  decoder.blocks.29.mlp.2.weight
decoder.blocks.29.mlp.2.weight 2 (1280, 5120)
model.decoder.layers.29.fc2.bias  ->  decoder.blocks.29.mlp.2.bias
decoder.blocks.29.mlp.2.bias 1 (1280,)
  Converting to float32
model.decoder.layers.29.final_layer_norm.weight  ->  decoder.blocks.29.mlp_ln.weight
decoder.blocks.29.mlp_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.29.final_layer_norm.bias  ->  decoder.blocks.29.mlp_ln.bias
decoder.blocks.29.mlp_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.30.self_attn.k_proj.weight  ->  decoder.blocks.30.attn.key.weight
decoder.blocks.30.attn.key.weight 2 (1280, 1280)
model.decoder.layers.30.self_attn.v_proj.weight  ->  decoder.blocks.30.attn.value.weight
decoder.blocks.30.attn.value.weight 2 (1280, 1280)
model.decoder.layers.30.self_attn.v_proj.bias  ->  decoder.blocks.30.attn.value.bias
decoder.blocks.30.attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.30.self_attn.q_proj.weight  ->  decoder.blocks.30.attn.query.weight
decoder.blocks.30.attn.query.weight 2 (1280, 1280)
model.decoder.layers.30.self_attn.q_proj.bias  ->  decoder.blocks.30.attn.query.bias
decoder.blocks.30.attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.30.self_attn.out_proj.weight  ->  decoder.blocks.30.attn.out.weight
decoder.blocks.30.attn.out.weight 2 (1280, 1280)
model.decoder.layers.30.self_attn.out_proj.bias  ->  decoder.blocks.30.attn.out.bias
decoder.blocks.30.attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.30.self_attn_layer_norm.weight  ->  decoder.blocks.30.attn_ln.weight
decoder.blocks.30.attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.30.self_attn_layer_norm.bias  ->  decoder.blocks.30.attn_ln.bias
decoder.blocks.30.attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.30.encoder_attn.k_proj.weight  ->  decoder.blocks.30.cross_attn.key.weight
decoder.blocks.30.cross_attn.key.weight 2 (1280, 1280)
model.decoder.layers.30.encoder_attn.v_proj.weight  ->  decoder.blocks.30.cross_attn.value.weight
decoder.blocks.30.cross_attn.value.weight 2 (1280, 1280)
model.decoder.layers.30.encoder_attn.v_proj.bias  ->  decoder.blocks.30.cross_attn.value.bias
decoder.blocks.30.cross_attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.30.encoder_attn.q_proj.weight  ->  decoder.blocks.30.cross_attn.query.weight
decoder.blocks.30.cross_attn.query.weight 2 (1280, 1280)
model.decoder.layers.30.encoder_attn.q_proj.bias  ->  decoder.blocks.30.cross_attn.query.bias
decoder.blocks.30.cross_attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.30.encoder_attn.out_proj.weight  ->  decoder.blocks.30.cross_attn.out.weight
decoder.blocks.30.cross_attn.out.weight 2 (1280, 1280)
model.decoder.layers.30.encoder_attn.out_proj.bias  ->  decoder.blocks.30.cross_attn.out.bias
decoder.blocks.30.cross_attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.30.encoder_attn_layer_norm.weight  ->  decoder.blocks.30.cross_attn_ln.weight
decoder.blocks.30.cross_attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.30.encoder_attn_layer_norm.bias  ->  decoder.blocks.30.cross_attn_ln.bias
decoder.blocks.30.cross_attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.30.fc1.weight  ->  decoder.blocks.30.mlp.0.weight
decoder.blocks.30.mlp.0.weight 2 (5120, 1280)
model.decoder.layers.30.fc1.bias  ->  decoder.blocks.30.mlp.0.bias
decoder.blocks.30.mlp.0.bias 1 (5120,)
  Converting to float32
model.decoder.layers.30.fc2.weight  ->  decoder.blocks.30.mlp.2.weight
decoder.blocks.30.mlp.2.weight 2 (1280, 5120)
model.decoder.layers.30.fc2.bias  ->  decoder.blocks.30.mlp.2.bias
decoder.blocks.30.mlp.2.bias 1 (1280,)
  Converting to float32
model.decoder.layers.30.final_layer_norm.weight  ->  decoder.blocks.30.mlp_ln.weight
decoder.blocks.30.mlp_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.30.final_layer_norm.bias  ->  decoder.blocks.30.mlp_ln.bias
decoder.blocks.30.mlp_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.31.self_attn.k_proj.weight  ->  decoder.blocks.31.attn.key.weight
decoder.blocks.31.attn.key.weight 2 (1280, 1280)
model.decoder.layers.31.self_attn.v_proj.weight  ->  decoder.blocks.31.attn.value.weight
decoder.blocks.31.attn.value.weight 2 (1280, 1280)
model.decoder.layers.31.self_attn.v_proj.bias  ->  decoder.blocks.31.attn.value.bias
decoder.blocks.31.attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.31.self_attn.q_proj.weight  ->  decoder.blocks.31.attn.query.weight
decoder.blocks.31.attn.query.weight 2 (1280, 1280)
model.decoder.layers.31.self_attn.q_proj.bias  ->  decoder.blocks.31.attn.query.bias
decoder.blocks.31.attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.31.self_attn.out_proj.weight  ->  decoder.blocks.31.attn.out.weight
decoder.blocks.31.attn.out.weight 2 (1280, 1280)
model.decoder.layers.31.self_attn.out_proj.bias  ->  decoder.blocks.31.attn.out.bias
decoder.blocks.31.attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.31.self_attn_layer_norm.weight  ->  decoder.blocks.31.attn_ln.weight
decoder.blocks.31.attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.31.self_attn_layer_norm.bias  ->  decoder.blocks.31.attn_ln.bias
decoder.blocks.31.attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.31.encoder_attn.k_proj.weight  ->  decoder.blocks.31.cross_attn.key.weight
decoder.blocks.31.cross_attn.key.weight 2 (1280, 1280)
model.decoder.layers.31.encoder_attn.v_proj.weight  ->  decoder.blocks.31.cross_attn.value.weight
decoder.blocks.31.cross_attn.value.weight 2 (1280, 1280)
model.decoder.layers.31.encoder_attn.v_proj.bias  ->  decoder.blocks.31.cross_attn.value.bias
decoder.blocks.31.cross_attn.value.bias 1 (1280,)
  Converting to float32
model.decoder.layers.31.encoder_attn.q_proj.weight  ->  decoder.blocks.31.cross_attn.query.weight
decoder.blocks.31.cross_attn.query.weight 2 (1280, 1280)
model.decoder.layers.31.encoder_attn.q_proj.bias  ->  decoder.blocks.31.cross_attn.query.bias
decoder.blocks.31.cross_attn.query.bias 1 (1280,)
  Converting to float32
model.decoder.layers.31.encoder_attn.out_proj.weight  ->  decoder.blocks.31.cross_attn.out.weight
decoder.blocks.31.cross_attn.out.weight 2 (1280, 1280)
model.decoder.layers.31.encoder_attn.out_proj.bias  ->  decoder.blocks.31.cross_attn.out.bias
decoder.blocks.31.cross_attn.out.bias 1 (1280,)
  Converting to float32
model.decoder.layers.31.encoder_attn_layer_norm.weight  ->  decoder.blocks.31.cross_attn_ln.weight
decoder.blocks.31.cross_attn_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.31.encoder_attn_layer_norm.bias  ->  decoder.blocks.31.cross_attn_ln.bias
decoder.blocks.31.cross_attn_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layers.31.fc1.weight  ->  decoder.blocks.31.mlp.0.weight
decoder.blocks.31.mlp.0.weight 2 (5120, 1280)
model.decoder.layers.31.fc1.bias  ->  decoder.blocks.31.mlp.0.bias
decoder.blocks.31.mlp.0.bias 1 (5120,)
  Converting to float32
model.decoder.layers.31.fc2.weight  ->  decoder.blocks.31.mlp.2.weight
decoder.blocks.31.mlp.2.weight 2 (1280, 5120)
model.decoder.layers.31.fc2.bias  ->  decoder.blocks.31.mlp.2.bias
decoder.blocks.31.mlp.2.bias 1 (1280,)
  Converting to float32
model.decoder.layers.31.final_layer_norm.weight  ->  decoder.blocks.31.mlp_ln.weight
decoder.blocks.31.mlp_ln.weight 1 (1280,)
  Converting to float32
model.decoder.layers.31.final_layer_norm.bias  ->  decoder.blocks.31.mlp_ln.bias
decoder.blocks.31.mlp_ln.bias 1 (1280,)
  Converting to float32
model.decoder.layer_norm.weight  ->  decoder.ln.weight
decoder.ln.weight 1 (1280,)
  Converting to float32
model.decoder.layer_norm.bias  ->  decoder.ln.bias
decoder.ln.bias 1 (1280,)
  Converting to float32
Skipping proj_out.weight
Done. Output file:  whisper-ggml-sme/ggml-model.bin

!wget https://media.globalrecordings.net/GOKit_MP3s_named/Saami%20North%20-%20The%20Two%20Roads.mp3 -O sample.mp3
!ffmpeg -i sample.mp3 -acodec pcm_s16le -ac 1 -ar 16000 sample.wav

--2024-03-03 18:09:42--  https://media.globalrecordings.net/GOKit_MP3s_named/Saami%20North%20-%20The%20Two%20Roads.mp3
Resolving media.globalrecordings.net (media.globalrecordings.net)... 35.208.248.145
Connecting to media.globalrecordings.net (media.globalrecordings.net)|35.208.248.145|:443... connected.
HTTP request sent, awaiting response... 200 OK
Length: 493511 (482K) [audio/mpeg]
Saving to: ‘sample.mp3’

sample.mp3          100%[===================>] 481.94K  3.00MB/s    in 0.2s    

2024-03-03 18:09:42 (3.00 MB/s) - ‘sample.mp3’ saved [493511/493511]

ffmpeg version 4.4.2-0ubuntu0.22.04.1 Copyright (c) 2000-2021 the FFmpeg developers
  built with gcc 11 (Ubuntu 11.2.0-19ubuntu1)
  configuration: --prefix=/usr --extra-version=0ubuntu0.22.04.1 --toolchain=hardened --libdir=/usr/lib/x86_64-linux-gnu --incdir=/usr/include/x86_64-linux-gnu --arch=amd64 --enable-gpl --disable-stripping --enable-gnutls --enable-ladspa --enable-libaom --enable-libass --enable-libbluray --enable-libbs2b --enable-libcaca --enable-libcdio --enable-libcodec2 --enable-libdav1d --enable-libflite --enable-libfontconfig --enable-libfreetype --enable-libfribidi --enable-libgme --enable-libgsm --enable-libjack --enable-libmp3lame --enable-libmysofa --enable-libopenjpeg --enable-libopenmpt --enable-libopus --enable-libpulse --enable-librabbitmq --enable-librubberband --enable-libshine --enable-libsnappy --enable-libsoxr --enable-libspeex --enable-libsrt --enable-libssh --enable-libtheora --enable-libtwolame --enable-libvidstab --enable-libvorbis --enable-libvpx --enable-libwebp --enable-libx265 --enable-libxml2 --enable-libxvid --enable-libzimg --enable-libzmq --enable-libzvbi --enable-lv2 --enable-omx --enable-openal --enable-opencl --enable-opengl --enable-sdl2 --enable-pocketsphinx --enable-librsvg --enable-libmfx --enable-libdc1394 --enable-libdrm --enable-libiec61883 --enable-chromaprint --enable-frei0r --enable-libx264 --enable-shared
  libavutil      56. 70.100 / 56. 70.100
  libavcodec     58.134.100 / 58.134.100
  libavformat    58. 76.100 / 58. 76.100
  libavdevice    58. 13.100 / 58. 13.100
  libavfilter     7.110.100 /  7.110.100
  libswscale      5.  9.100 /  5.  9.100
  libswresample   3.  9.100 /  3.  9.100
  libpostproc    55.  9.100 / 55.  9.100
Input #0, mp3, from 'sample.mp3':
  Metadata:
    title           : Saami, North Picture 22: The Two Roads The Two Roads
    artist          : GRN Language Samples
    album           : Saami, North SME
    composer        : Saami, North
    genre           : GRN Language Sample
    encoder         : Lavf58.76.100
    track           : 1
    copyright       : 2010 GRN
    comment         : https://globalrecordings.net/en/language/3475
    grnprog         : A63258
    grnlang         : 3475
    grnprepared     : 20240113
    date            : 2010
  Duration: 00:01:44.67, start: 0.050111, bitrate: 37 kb/s
  Stream #0:0: Audio: mp3, 22050 Hz, mono, fltp, 32 kb/s
  Stream #0:1: Video: mjpeg (Baseline), yuvj420p(pc, bt470bg/unknown/unknown), 300x300 [SAR 300:300 DAR 1:1], 90k tbr, 90k tbn, 90k tbc (attached pic)
    Metadata:
      title           : gn-22
      comment         : Cover (front)
Stream mapping:
  Stream #0:0 -> #0:0 (mp3 (mp3float) -> pcm_s16le (native))
Press [q] to stop, [?] for help
Output #0, wav, to 'sample.wav':
  Metadata:
    INAM            : Saami, North Picture 22: The Two Roads The Two Roads
    IART            : GRN Language Samples
    IPRD            : Saami, North SME
    composer        : Saami, North
    IGNR            : GRN Language Sample
    ICRD            : 2010
    IPRT            : 1
    ICOP            : 2010 GRN
    ICMT            : https://globalrecordings.net/en/language/3475
    grnprog         : A63258
    grnlang         : 3475
    grnprepared     : 20240113
    ISFT            : Lavf58.76.100
  Stream #0:0: Audio: pcm_s16le ([1][0][0][0] / 0x0001), 16000 Hz, mono, s16, 256 kb/s
    Metadata:
      encoder         : Lavc58.134.100 pcm_s16le
size=    3269kB time=00:01:44.61 bitrate= 256.0kbits/s speed= 722x    
video:0kB audio:3269kB subtitle:0kB other streams:0kB global headers:0kB muxing overhead: 0.009559%

!./whisper.cpp/main  -m whisper-ggml-sme/ggml-model.bin -f sample.wav

whisper_init_from_file_with_params_no_state: loading model from 'whisper-ggml-sme/ggml-model.bin'
whisper_model_load: loading model
whisper_model_load: n_vocab       = 51865
whisper_model_load: n_audio_ctx   = 1500
whisper_model_load: n_audio_state = 1280
whisper_model_load: n_audio_head  = 20
whisper_model_load: n_audio_layer = 32
whisper_model_load: n_text_ctx    = 448
whisper_model_load: n_text_state  = 1280
whisper_model_load: n_text_head   = 20
whisper_model_load: n_text_layer  = 32
whisper_model_load: n_mels        = 80
whisper_model_load: ftype         = 1
whisper_model_load: qntvr         = 0
whisper_model_load: type          = 5 (large)
whisper_model_load: adding 1608 extra tokens
whisper_model_load: n_langs       = 99
whisper_model_load:      CPU total size =  3093.99 MB
whisper_model_load: model size    = 3093.99 MB
whisper_init_state: kv self size  =  220.20 MB
whisper_init_state: kv cross size =  245.76 MB
whisper_init_state: compute buffer (conv)   =   34.82 MB
whisper_init_state: compute buffer (encode) =  926.66 MB
whisper_init_state: compute buffer (cross)  =    9.38 MB
whisper_init_state: compute buffer (decode) =  209.26 MB

system_info: n_threads = 2 / 2 | AVX = 1 | AVX2 = 1 | AVX512 = 0 | FMA = 1 | NEON = 0 | ARM_FMA = 0 | METAL = 0 | F16C = 1 | FP16_VA = 0 | WASM_SIMD = 0 | BLAS = 0 | SSE3 = 1 | SSSE3 = 1 | VSX = 0 | CUDA = 0 | COREML = 0 | OPENVINO = 0 | 

main: processing 'sample.wav' (1673814 samples, 104.6 sec), 2 threads, 1 processors, 5 beams + best of 5, lang = en, task = transcribe, timestamps = 1 ...