Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

ROCm error: ggml-cuda.cu:6246: invalid device function #3320

Closed
Displacer opened this issue Sep 23, 2023 · 14 comments · Fixed by #3853
Closed

ROCm error: ggml-cuda.cu:6246: invalid device function #3320

Displacer opened this issue Sep 23, 2023 · 14 comments · Fixed by #3853

Comments

@Displacer
Copy link

Running inference with
./main -ngl 1 -m llama-2-13b-chat.Q5_K_M.gguf -p "write hello"
an error appears:
Log start
main: build = 1268 (bedb92b)
main: built with AMD clang version 17.0.0 (https://github.com/RadeonOpenCompute/llvm-project roc-5.7.0 23352 d1e13c532a947d0cbfc94759c00dcf152294aa13) for x86_64-unknown-linux-gnu
main: seed = 1695481826
ggml_init_cublas: found 1 ROCm devices:
Device 0: , compute capability 11.0

... ...

CUDA error 98 at /code/llama.cpp/ggml-cuda.cu:6246: invalid device function

running with Ubuntu 22, rocm 5.7, AMD Radeon 7900 XTX in docker environment

with run command:

docker run -it --network=host --device=/dev/kfd --device=/dev/dri --ipc=host --shm-size 24G --group-add video --cap-add=SYS_PTRACE --security-opt seccomp=unconfined
@staviq
Copy link
Contributor

staviq commented Sep 23, 2023

Please post the full build command you used.

And check if you can run main without docker.

@Dasug
Copy link

Dasug commented Sep 24, 2023

I managed to trigger the same error when building with cmake the following way:

mkdir build
cd build
CC=/opt/rocm/llvm/bin/clang CXX=/opt/rocm/llvm/bin/clang++ cmake .. -DLLAMA_HIPBLAS=ON
cmake --build .

I also have the following environment variables set:

export HSA_OVERRIDE_GFX_VERSION=11.0.0
export HCC_AMDGPU_TARGET=gfx1100

calling ./main -ngl 43 -m ../../models/athena-v1.Q5_K_M.gguf -p "write hello" then gives me:

Log start
main: warning: changing RoPE frequency base to 0 (default 10000.0)
main: warning: scaling RoPE frequency by 0 (default 1.0)
main: build = 1270 (c091cdf)
main: built with clang version 16.0.0 for x86_64-pc-linux-gnu
main: seed  = 1695514502
ggml_init_cublas: found 1 ROCm devices:
  Device 0: AMD Radeon RX 7900 XTX, compute capability 11.0
llama_model_loader: loaded meta data with 19 key-value pairs and 363 tensors from ../../models/athena-v1.Q5_K_M.gguf (version GGUF V2 (latest))
llama_model_loader: - tensor    0:                token_embd.weight q5_K     [  5120, 32000,     1,     1 ]
llama_model_loader: - tensor    1:           blk.0.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor    2:            blk.0.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor    3:            blk.0.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor    4:              blk.0.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor    5:            blk.0.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor    6:              blk.0.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor    7:         blk.0.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor    8:              blk.0.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor    9:              blk.0.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   10:           blk.1.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   11:            blk.1.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor   12:            blk.1.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   13:              blk.1.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   14:            blk.1.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   15:              blk.1.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   16:         blk.1.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   17:              blk.1.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   18:              blk.1.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   19:            blk.2.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   20:              blk.2.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   21:              blk.2.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   22:         blk.2.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   23:              blk.2.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   24:              blk.2.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   25:           blk.2.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   26:            blk.2.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor   27:            blk.2.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   28:           blk.3.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   29:            blk.3.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor   30:            blk.3.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   31:              blk.3.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   32:            blk.3.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   33:              blk.3.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   34:         blk.3.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   35:              blk.3.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   36:              blk.3.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   37:           blk.4.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   38:            blk.4.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor   39:            blk.4.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   40:              blk.4.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   41:            blk.4.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   42:              blk.4.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   43:         blk.4.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   44:              blk.4.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   45:              blk.4.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   46:           blk.5.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   47:            blk.5.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor   48:            blk.5.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   49:              blk.5.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   50:            blk.5.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   51:              blk.5.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   52:         blk.5.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   53:              blk.5.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   54:              blk.5.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   55:           blk.6.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   56:            blk.6.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor   57:            blk.6.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   58:              blk.6.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   59:            blk.6.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   60:              blk.6.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   61:         blk.6.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   62:              blk.6.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   63:              blk.6.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   64:           blk.7.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   65:            blk.7.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor   66:            blk.7.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   67:              blk.7.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   68:            blk.7.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   69:              blk.7.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   70:         blk.7.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   71:              blk.7.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   72:              blk.7.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   73:           blk.8.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   74:            blk.8.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor   75:            blk.8.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   76:              blk.8.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   77:            blk.8.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   78:              blk.8.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   79:         blk.8.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   80:              blk.8.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   81:              blk.8.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   82:              blk.9.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   83:              blk.9.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   84:              blk.9.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   85:          blk.10.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   86:           blk.10.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor   87:           blk.10.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   88:             blk.10.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   89:           blk.10.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   90:             blk.10.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   91:        blk.10.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   92:             blk.10.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   93:             blk.10.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   94:          blk.11.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   95:           blk.11.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor   96:           blk.11.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   97:             blk.11.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   98:           blk.11.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   99:             blk.11.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  100:        blk.11.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  101:             blk.11.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  102:             blk.11.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  103:             blk.12.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  104:        blk.12.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  105:             blk.12.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  106:             blk.12.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  107:           blk.9.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  108:            blk.9.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  109:            blk.9.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  110:              blk.9.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  111:            blk.9.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  112:         blk.9.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  113:          blk.12.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  114:           blk.12.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  115:           blk.12.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  116:             blk.12.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  117:           blk.12.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  118:          blk.13.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  119:           blk.13.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  120:           blk.13.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  121:             blk.13.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  122:           blk.13.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  123:             blk.13.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  124:        blk.13.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  125:             blk.13.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  126:             blk.13.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  127:          blk.14.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  128:           blk.14.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  129:           blk.14.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  130:             blk.14.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  131:           blk.14.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  132:             blk.14.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  133:        blk.14.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  134:             blk.14.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  135:             blk.14.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  136:           blk.15.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  137:             blk.15.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  138:        blk.15.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  139:             blk.15.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  140:             blk.15.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  141:          blk.15.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  142:           blk.15.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  143:             blk.15.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  144:           blk.15.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  145:          blk.16.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  146:           blk.16.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  147:           blk.16.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  148:             blk.16.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  149:           blk.16.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  150:             blk.16.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  151:        blk.16.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  152:             blk.16.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  153:             blk.16.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  154:          blk.17.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  155:           blk.17.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  156:           blk.17.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  157:             blk.17.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  158:           blk.17.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  159:             blk.17.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  160:        blk.17.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  161:             blk.17.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  162:             blk.17.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  163:           blk.18.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  164:             blk.18.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  165:             blk.18.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  166:        blk.18.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  167:             blk.18.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  168:             blk.18.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  169:          blk.18.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  170:           blk.18.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  171:           blk.18.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  172:          blk.19.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  173:           blk.19.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  174:           blk.19.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  175:             blk.19.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  176:           blk.19.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  177:             blk.19.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  178:        blk.19.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  179:             blk.19.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  180:             blk.19.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  181:          blk.20.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  182:           blk.20.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  183:           blk.20.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  184:             blk.20.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  185:           blk.20.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  186:             blk.20.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  187:        blk.20.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  188:             blk.20.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  189:             blk.20.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  190:          blk.21.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  191:           blk.21.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  192:           blk.21.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  193:             blk.21.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  194:           blk.21.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  195:             blk.21.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  196:        blk.21.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  197:             blk.21.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  198:             blk.21.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  199:          blk.22.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  200:           blk.22.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  201:           blk.22.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  202:             blk.22.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  203:           blk.22.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  204:             blk.22.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  205:        blk.22.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  206:             blk.22.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  207:             blk.22.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  208:          blk.23.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  209:           blk.23.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  210:           blk.23.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  211:             blk.23.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  212:           blk.23.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  213:             blk.23.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  214:        blk.23.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  215:             blk.23.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  216:             blk.23.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  217:          blk.24.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  218:           blk.24.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  219:           blk.24.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  220:             blk.24.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  221:           blk.24.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  222:             blk.24.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  223:        blk.24.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  224:             blk.24.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  225:             blk.24.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  226:             blk.25.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  227:             blk.25.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  228:             blk.25.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  229:          blk.25.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  230:           blk.25.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  231:           blk.25.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  232:             blk.25.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  233:           blk.25.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  234:        blk.25.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  235:          blk.26.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  236:           blk.26.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  237:           blk.26.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  238:             blk.26.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  239:           blk.26.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  240:             blk.26.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  241:        blk.26.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  242:             blk.26.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  243:             blk.26.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  244:          blk.27.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  245:           blk.27.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  246:           blk.27.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  247:             blk.27.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  248:           blk.27.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  249:             blk.27.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  250:        blk.27.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  251:             blk.27.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  252:             blk.27.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  253:             blk.28.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  254:        blk.28.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  255:             blk.28.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  256:             blk.28.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  257:          blk.28.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  258:           blk.28.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  259:           blk.28.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  260:             blk.28.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  261:           blk.28.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  262:          blk.29.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  263:           blk.29.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  264:           blk.29.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  265:             blk.29.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  266:           blk.29.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  267:             blk.29.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  268:        blk.29.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  269:             blk.29.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  270:             blk.29.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  271:          blk.30.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  272:           blk.30.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  273:           blk.30.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  274:             blk.30.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  275:           blk.30.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  276:             blk.30.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  277:        blk.30.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  278:             blk.30.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  279:             blk.30.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  280:           blk.31.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  281:             blk.31.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  282:        blk.31.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  283:             blk.31.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  284:             blk.31.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  285:          blk.31.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  286:           blk.31.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  287:             blk.31.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  288:           blk.31.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  289:          blk.32.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  290:           blk.32.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  291:           blk.32.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  292:             blk.32.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  293:           blk.32.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  294:             blk.32.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  295:        blk.32.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  296:             blk.32.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  297:             blk.32.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  298:          blk.33.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  299:           blk.33.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  300:           blk.33.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  301:             blk.33.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  302:           blk.33.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  303:             blk.33.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  304:        blk.33.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  305:             blk.33.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  306:             blk.33.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  307:           blk.34.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  308:             blk.34.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  309:             blk.34.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  310:        blk.34.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  311:             blk.34.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  312:             blk.34.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  313:          blk.34.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  314:           blk.34.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  315:           blk.34.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  316:          blk.35.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  317:           blk.35.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  318:           blk.35.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  319:             blk.35.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  320:           blk.35.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  321:             blk.35.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  322:        blk.35.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  323:             blk.35.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  324:             blk.35.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  325:          blk.36.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  326:           blk.36.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  327:           blk.36.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  328:             blk.36.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  329:           blk.36.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  330:             blk.36.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  331:        blk.36.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  332:             blk.36.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  333:             blk.36.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  334:          blk.37.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  335:           blk.37.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  336:           blk.37.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  337:             blk.37.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  338:           blk.37.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  339:             blk.37.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  340:        blk.37.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  341:             blk.37.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  342:             blk.37.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  343:                    output.weight q6_K     [  5120, 32000,     1,     1 ]
llama_model_loader: - tensor  344:          blk.38.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  345:           blk.38.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  346:           blk.38.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  347:             blk.38.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  348:           blk.38.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  349:             blk.38.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  350:        blk.38.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  351:             blk.38.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  352:             blk.38.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  353:          blk.39.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  354:           blk.39.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  355:           blk.39.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  356:             blk.39.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  357:           blk.39.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  358:             blk.39.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  359:        blk.39.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  360:             blk.39.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  361:             blk.39.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  362:               output_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - kv   0:                       general.architecture str     
llama_model_loader: - kv   1:                               general.name str     
llama_model_loader: - kv   2:                       llama.context_length u32     
llama_model_loader: - kv   3:                     llama.embedding_length u32     
llama_model_loader: - kv   4:                          llama.block_count u32     
llama_model_loader: - kv   5:                  llama.feed_forward_length u32     
llama_model_loader: - kv   6:                 llama.rope.dimension_count u32     
llama_model_loader: - kv   7:                 llama.attention.head_count u32     
llama_model_loader: - kv   8:              llama.attention.head_count_kv u32     
llama_model_loader: - kv   9:     llama.attention.layer_norm_rms_epsilon f32     
llama_model_loader: - kv  10:                          general.file_type u32     
llama_model_loader: - kv  11:                       tokenizer.ggml.model str     
llama_model_loader: - kv  12:                      tokenizer.ggml.tokens arr     
llama_model_loader: - kv  13:                      tokenizer.ggml.scores arr     
llama_model_loader: - kv  14:                  tokenizer.ggml.token_type arr     
llama_model_loader: - kv  15:                tokenizer.ggml.bos_token_id u32     
llama_model_loader: - kv  16:                tokenizer.ggml.eos_token_id u32     
llama_model_loader: - kv  17:            tokenizer.ggml.unknown_token_id u32     
llama_model_loader: - kv  18:               general.quantization_version u32     
llama_model_loader: - type  f32:   81 tensors
llama_model_loader: - type q5_K:  241 tensors
llama_model_loader: - type q6_K:   41 tensors
llm_load_print_meta: format         = GGUF V2 (latest)
llm_load_print_meta: arch           = llama
llm_load_print_meta: vocab type     = SPM
llm_load_print_meta: n_vocab        = 32000
llm_load_print_meta: n_merges       = 0
llm_load_print_meta: n_ctx_train    = 4096
llm_load_print_meta: n_ctx          = 512
llm_load_print_meta: n_embd         = 5120
llm_load_print_meta: n_head         = 40
llm_load_print_meta: n_head_kv      = 40
llm_load_print_meta: n_layer        = 40
llm_load_print_meta: n_rot          = 128
llm_load_print_meta: n_gqa          = 1
llm_load_print_meta: f_norm_eps     = 0,0e+00
llm_load_print_meta: f_norm_rms_eps = 1,0e-05
llm_load_print_meta: n_ff           = 13824
llm_load_print_meta: freq_base      = 10000,0
llm_load_print_meta: freq_scale     = 1
llm_load_print_meta: model type     = 13B
llm_load_print_meta: model ftype    = mostly Q5_K - Medium
llm_load_print_meta: model params   = 13,02 B
llm_load_print_meta: model size     = 8,60 GiB (5,67 BPW) 
llm_load_print_meta: general.name   = LLaMA v2
llm_load_print_meta: BOS token = 1 '<s>'
llm_load_print_meta: EOS token = 2 '</s>'
llm_load_print_meta: UNK token = 0 '<unk>'
llm_load_print_meta: LF token  = 13 '<0x0A>'
llm_load_tensors: ggml ctx size =    0,12 MB
llm_load_tensors: using ROCm for GPU acceleration
llm_load_tensors: mem required  =  107,54 MB (+  400,00 MB per state)
llm_load_tensors: offloading 40 repeating layers to GPU
llm_load_tensors: offloading non-repeating layers to GPU
llm_load_tensors: offloading v cache to GPU
llm_load_tensors: offloading k cache to GPU
llm_load_tensors: offloaded 43/43 layers to GPU
llm_load_tensors: VRAM used: 9095 MB
....................................................................................................
llama_new_context_with_model: kv self size  =  400,00 MB
llama_new_context_with_model: compute buffer total size =   75,47 MB
llama_new_context_with_model: VRAM scratch buffer: 74,00 MB

CUDA error 98 at /run/media/firestorm-stormfront/user/dasug/repositories/llama.cpp/ggml-cuda.cu:6246: invalid device function
current device: 0

(using a different model here since I only have this one at hand currently, sorry! If necessary I can re-test with a supported model later.)

Surprisingly the error does not occur when I build it with make (make LLAMA_HIPBLAS=1):

./main -ngl 43 -m models/athena-v1.Q5_K_M.gguf -p "write hello"
Log start
main: warning: changing RoPE frequency base to 0 (default 10000.0)
main: warning: scaling RoPE frequency by 0 (default 1.0)
main: build = 1270 (c091cdf)
main: built with cc (GCC) 13.2.1 20230801 for x86_64-pc-linux-gnu
main: seed  = 1695515082
ggml_init_cublas: found 1 ROCm devices:
  Device 0: AMD Radeon RX 7900 XTX, compute capability 11.0
llama_model_loader: loaded meta data with 19 key-value pairs and 363 tensors from models/athena-v1.Q5_K_M.gguf (version GGUF V2 (latest))
llama_model_loader: - tensor    0:                token_embd.weight q5_K     [  5120, 32000,     1,     1 ]
llama_model_loader: - tensor    1:           blk.0.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor    2:            blk.0.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor    3:            blk.0.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor    4:              blk.0.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor    5:            blk.0.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor    6:              blk.0.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor    7:         blk.0.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor    8:              blk.0.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor    9:              blk.0.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   10:           blk.1.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   11:            blk.1.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor   12:            blk.1.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   13:              blk.1.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   14:            blk.1.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   15:              blk.1.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   16:         blk.1.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   17:              blk.1.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   18:              blk.1.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   19:            blk.2.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   20:              blk.2.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   21:              blk.2.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   22:         blk.2.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   23:              blk.2.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   24:              blk.2.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   25:           blk.2.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   26:            blk.2.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor   27:            blk.2.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   28:           blk.3.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   29:            blk.3.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor   30:            blk.3.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   31:              blk.3.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   32:            blk.3.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   33:              blk.3.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   34:         blk.3.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   35:              blk.3.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   36:              blk.3.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   37:           blk.4.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   38:            blk.4.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor   39:            blk.4.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   40:              blk.4.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   41:            blk.4.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   42:              blk.4.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   43:         blk.4.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   44:              blk.4.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   45:              blk.4.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   46:           blk.5.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   47:            blk.5.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor   48:            blk.5.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   49:              blk.5.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   50:            blk.5.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   51:              blk.5.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   52:         blk.5.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   53:              blk.5.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   54:              blk.5.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   55:           blk.6.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   56:            blk.6.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor   57:            blk.6.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   58:              blk.6.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   59:            blk.6.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   60:              blk.6.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   61:         blk.6.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   62:              blk.6.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   63:              blk.6.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   64:           blk.7.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   65:            blk.7.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor   66:            blk.7.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   67:              blk.7.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   68:            blk.7.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   69:              blk.7.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   70:         blk.7.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   71:              blk.7.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   72:              blk.7.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   73:           blk.8.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   74:            blk.8.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor   75:            blk.8.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   76:              blk.8.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   77:            blk.8.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   78:              blk.8.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   79:         blk.8.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   80:              blk.8.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   81:              blk.8.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   82:              blk.9.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   83:              blk.9.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   84:              blk.9.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   85:          blk.10.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   86:           blk.10.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor   87:           blk.10.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   88:             blk.10.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   89:           blk.10.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   90:             blk.10.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   91:        blk.10.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   92:             blk.10.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   93:             blk.10.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor   94:          blk.11.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   95:           blk.11.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor   96:           blk.11.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   97:             blk.11.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor   98:           blk.11.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor   99:             blk.11.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  100:        blk.11.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  101:             blk.11.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  102:             blk.11.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  103:             blk.12.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  104:        blk.12.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  105:             blk.12.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  106:             blk.12.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  107:           blk.9.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  108:            blk.9.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  109:            blk.9.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  110:              blk.9.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  111:            blk.9.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  112:         blk.9.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  113:          blk.12.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  114:           blk.12.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  115:           blk.12.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  116:             blk.12.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  117:           blk.12.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  118:          blk.13.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  119:           blk.13.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  120:           blk.13.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  121:             blk.13.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  122:           blk.13.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  123:             blk.13.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  124:        blk.13.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  125:             blk.13.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  126:             blk.13.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  127:          blk.14.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  128:           blk.14.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  129:           blk.14.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  130:             blk.14.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  131:           blk.14.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  132:             blk.14.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  133:        blk.14.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  134:             blk.14.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  135:             blk.14.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  136:           blk.15.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  137:             blk.15.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  138:        blk.15.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  139:             blk.15.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  140:             blk.15.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  141:          blk.15.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  142:           blk.15.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  143:             blk.15.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  144:           blk.15.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  145:          blk.16.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  146:           blk.16.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  147:           blk.16.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  148:             blk.16.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  149:           blk.16.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  150:             blk.16.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  151:        blk.16.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  152:             blk.16.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  153:             blk.16.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  154:          blk.17.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  155:           blk.17.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  156:           blk.17.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  157:             blk.17.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  158:           blk.17.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  159:             blk.17.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  160:        blk.17.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  161:             blk.17.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  162:             blk.17.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  163:           blk.18.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  164:             blk.18.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  165:             blk.18.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  166:        blk.18.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  167:             blk.18.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  168:             blk.18.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  169:          blk.18.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  170:           blk.18.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  171:           blk.18.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  172:          blk.19.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  173:           blk.19.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  174:           blk.19.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  175:             blk.19.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  176:           blk.19.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  177:             blk.19.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  178:        blk.19.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  179:             blk.19.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  180:             blk.19.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  181:          blk.20.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  182:           blk.20.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  183:           blk.20.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  184:             blk.20.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  185:           blk.20.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  186:             blk.20.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  187:        blk.20.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  188:             blk.20.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  189:             blk.20.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  190:          blk.21.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  191:           blk.21.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  192:           blk.21.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  193:             blk.21.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  194:           blk.21.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  195:             blk.21.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  196:        blk.21.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  197:             blk.21.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  198:             blk.21.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  199:          blk.22.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  200:           blk.22.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  201:           blk.22.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  202:             blk.22.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  203:           blk.22.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  204:             blk.22.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  205:        blk.22.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  206:             blk.22.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  207:             blk.22.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  208:          blk.23.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  209:           blk.23.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  210:           blk.23.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  211:             blk.23.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  212:           blk.23.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  213:             blk.23.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  214:        blk.23.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  215:             blk.23.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  216:             blk.23.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  217:          blk.24.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  218:           blk.24.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  219:           blk.24.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  220:             blk.24.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  221:           blk.24.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  222:             blk.24.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  223:        blk.24.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  224:             blk.24.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  225:             blk.24.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  226:             blk.25.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  227:             blk.25.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  228:             blk.25.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  229:          blk.25.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  230:           blk.25.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  231:           blk.25.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  232:             blk.25.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  233:           blk.25.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  234:        blk.25.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  235:          blk.26.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  236:           blk.26.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  237:           blk.26.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  238:             blk.26.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  239:           blk.26.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  240:             blk.26.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  241:        blk.26.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  242:             blk.26.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  243:             blk.26.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  244:          blk.27.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  245:           blk.27.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  246:           blk.27.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  247:             blk.27.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  248:           blk.27.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  249:             blk.27.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  250:        blk.27.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  251:             blk.27.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  252:             blk.27.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  253:             blk.28.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  254:        blk.28.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  255:             blk.28.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  256:             blk.28.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  257:          blk.28.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  258:           blk.28.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  259:           blk.28.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  260:             blk.28.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  261:           blk.28.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  262:          blk.29.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  263:           blk.29.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  264:           blk.29.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  265:             blk.29.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  266:           blk.29.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  267:             blk.29.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  268:        blk.29.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  269:             blk.29.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  270:             blk.29.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  271:          blk.30.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  272:           blk.30.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  273:           blk.30.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  274:             blk.30.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  275:           blk.30.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  276:             blk.30.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  277:        blk.30.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  278:             blk.30.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  279:             blk.30.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  280:           blk.31.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  281:             blk.31.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  282:        blk.31.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  283:             blk.31.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  284:             blk.31.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  285:          blk.31.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  286:           blk.31.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  287:             blk.31.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  288:           blk.31.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  289:          blk.32.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  290:           blk.32.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  291:           blk.32.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  292:             blk.32.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  293:           blk.32.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  294:             blk.32.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  295:        blk.32.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  296:             blk.32.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  297:             blk.32.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  298:          blk.33.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  299:           blk.33.ffn_down.weight q5_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  300:           blk.33.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  301:             blk.33.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  302:           blk.33.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  303:             blk.33.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  304:        blk.33.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  305:             blk.33.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  306:             blk.33.attn_v.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  307:           blk.34.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  308:             blk.34.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  309:             blk.34.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  310:        blk.34.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  311:             blk.34.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  312:             blk.34.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  313:          blk.34.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  314:           blk.34.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  315:           blk.34.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  316:          blk.35.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  317:           blk.35.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  318:           blk.35.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  319:             blk.35.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  320:           blk.35.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  321:             blk.35.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  322:        blk.35.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  323:             blk.35.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  324:             blk.35.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  325:          blk.36.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  326:           blk.36.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  327:           blk.36.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  328:             blk.36.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  329:           blk.36.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  330:             blk.36.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  331:        blk.36.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  332:             blk.36.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  333:             blk.36.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  334:          blk.37.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  335:           blk.37.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  336:           blk.37.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  337:             blk.37.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  338:           blk.37.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  339:             blk.37.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  340:        blk.37.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  341:             blk.37.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  342:             blk.37.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  343:                    output.weight q6_K     [  5120, 32000,     1,     1 ]
llama_model_loader: - tensor  344:          blk.38.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  345:           blk.38.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  346:           blk.38.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  347:             blk.38.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  348:           blk.38.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  349:             blk.38.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  350:        blk.38.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  351:             blk.38.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  352:             blk.38.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  353:          blk.39.attn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  354:           blk.39.ffn_down.weight q6_K     [ 13824,  5120,     1,     1 ]
llama_model_loader: - tensor  355:           blk.39.ffn_gate.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  356:             blk.39.ffn_up.weight q5_K     [  5120, 13824,     1,     1 ]
llama_model_loader: - tensor  357:           blk.39.ffn_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - tensor  358:             blk.39.attn_k.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  359:        blk.39.attn_output.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  360:             blk.39.attn_q.weight q5_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  361:             blk.39.attn_v.weight q6_K     [  5120,  5120,     1,     1 ]
llama_model_loader: - tensor  362:               output_norm.weight f32      [  5120,     1,     1,     1 ]
llama_model_loader: - kv   0:                       general.architecture str     
llama_model_loader: - kv   1:                               general.name str     
llama_model_loader: - kv   2:                       llama.context_length u32     
llama_model_loader: - kv   3:                     llama.embedding_length u32     
llama_model_loader: - kv   4:                          llama.block_count u32     
llama_model_loader: - kv   5:                  llama.feed_forward_length u32     
llama_model_loader: - kv   6:                 llama.rope.dimension_count u32     
llama_model_loader: - kv   7:                 llama.attention.head_count u32     
llama_model_loader: - kv   8:              llama.attention.head_count_kv u32     
llama_model_loader: - kv   9:     llama.attention.layer_norm_rms_epsilon f32     
llama_model_loader: - kv  10:                          general.file_type u32     
llama_model_loader: - kv  11:                       tokenizer.ggml.model str     
llama_model_loader: - kv  12:                      tokenizer.ggml.tokens arr     
llama_model_loader: - kv  13:                      tokenizer.ggml.scores arr     
llama_model_loader: - kv  14:                  tokenizer.ggml.token_type arr     
llama_model_loader: - kv  15:                tokenizer.ggml.bos_token_id u32     
llama_model_loader: - kv  16:                tokenizer.ggml.eos_token_id u32     
llama_model_loader: - kv  17:            tokenizer.ggml.unknown_token_id u32     
llama_model_loader: - kv  18:               general.quantization_version u32     
llama_model_loader: - type  f32:   81 tensors
llama_model_loader: - type q5_K:  241 tensors
llama_model_loader: - type q6_K:   41 tensors
llm_load_print_meta: format         = GGUF V2 (latest)
llm_load_print_meta: arch           = llama
llm_load_print_meta: vocab type     = SPM
llm_load_print_meta: n_vocab        = 32000
llm_load_print_meta: n_merges       = 0
llm_load_print_meta: n_ctx_train    = 4096
llm_load_print_meta: n_ctx          = 512
llm_load_print_meta: n_embd         = 5120
llm_load_print_meta: n_head         = 40
llm_load_print_meta: n_head_kv      = 40
llm_load_print_meta: n_layer        = 40
llm_load_print_meta: n_rot          = 128
llm_load_print_meta: n_gqa          = 1
llm_load_print_meta: f_norm_eps     = 0,0e+00
llm_load_print_meta: f_norm_rms_eps = 1,0e-05
llm_load_print_meta: n_ff           = 13824
llm_load_print_meta: freq_base      = 10000,0
llm_load_print_meta: freq_scale     = 1
llm_load_print_meta: model type     = 13B
llm_load_print_meta: model ftype    = mostly Q5_K - Medium
llm_load_print_meta: model params   = 13,02 B
llm_load_print_meta: model size     = 8,60 GiB (5,67 BPW) 
llm_load_print_meta: general.name   = LLaMA v2
llm_load_print_meta: BOS token = 1 '<s>'
llm_load_print_meta: EOS token = 2 '</s>'
llm_load_print_meta: UNK token = 0 '<unk>'
llm_load_print_meta: LF token  = 13 '<0x0A>'
llm_load_tensors: ggml ctx size =    0,12 MB
llm_load_tensors: using ROCm for GPU acceleration
llm_load_tensors: mem required  =  107,54 MB (+  400,00 MB per state)
llm_load_tensors: offloading 40 repeating layers to GPU
llm_load_tensors: offloading non-repeating layers to GPU
llm_load_tensors: offloading v cache to GPU
llm_load_tensors: offloading k cache to GPU
llm_load_tensors: offloaded 43/43 layers to GPU
llm_load_tensors: VRAM used: 9095 MB
....................................................................................................
llama_new_context_with_model: kv self size  =  400,00 MB
llama_new_context_with_model: compute buffer total size =   75,47 MB
llama_new_context_with_model: VRAM scratch buffer: 74,00 MB

system_info: n_threads = 12 / 24 | AVX = 1 | AVX2 = 1 | AVX512 = 0 | AVX512_VBMI = 0 | AVX512_VNNI = 0 | FMA = 1 | NEON = 0 | ARM_FMA = 0 | F16C = 1 | FP16_VA = 0 | WASM_SIMD = 0 | BLAS = 1 | SSE3 = 1 | SSSE3 = 1 | VSX = 0 | 
sampling: repeat_last_n = 64, repeat_penalty = 1,100000, presence_penalty = 0,000000, frequency_penalty = 0,000000, top_k = 40, tfs_z = 1,000000, top_p = 0,950000, typical_p = 1,000000, temp = 0,800000, mirostat = 0, mirostat_lr = 0,100000, mirostat_ent = 5,000000
generate: n_ctx = 512, n_batch = 512, n_predict = -1, n_keep = 0


 write hello world
```rust
fn main() {
    println!("Hello, World!");
}
```
 [end of text]

llama_print_timings:        load time =  4493,92 ms
llama_print_timings:      sample time =     8,88 ms /    23 runs   (    0,39 ms per token,  2589,51 tokens per second)
llama_print_timings: prompt eval time =   135,28 ms /     3 tokens (   45,09 ms per token,    22,18 tokens per second)
llama_print_timings:        eval time =   666,13 ms /    22 runs   (   30,28 ms per token,    33,03 tokens per second)
llama_print_timings:       total time =   818,67 ms
Log end

@staviq
Copy link
Contributor

staviq commented Sep 24, 2023

Can you try cmake without forcing CC/CXX ?

If make defaulted to gcc and worked, the problem might be specific to clang

@Dasug
Copy link

Dasug commented Sep 24, 2023

Building from cmake without forcing CC/CXX fails entirely since cmake is trying to build with my regular system gcc (which is why I assume this project's README.md forces CC/CXX as well when using cmake with hipBLAS):

[  1%] Built target BUILD_INFO
[  2%] Building CXX object CMakeFiles/ggml-rocm.dir/ggml-cuda.cu.o
c++: error: unrecognized command-line option ‘--offload-arch=gfx900’
c++: error: unrecognized command-line option ‘--offload-arch=gfx906’
c++: error: unrecognized command-line option ‘--offload-arch=gfx908’
c++: error: unrecognized command-line option ‘--offload-arch=gfx90a’
c++: error: unrecognized command-line option ‘--offload-arch=gfx1030’
make[2]: *** [CMakeFiles/ggml-rocm.dir/build.make:76: CMakeFiles/ggml-rocm.dir/ggml-cuda.cu.o] Error 1
make[1]: *** [CMakeFiles/Makefile2:538: CMakeFiles/ggml-rocm.dir/all] Error 2
make: *** [Makefile:146: all] Error 2

Looking over the Makefile, it seems like it specifically compiles ggml-cuda with clang included within ROCm if LLAMA_HIPBLAS is set and for the rest it uses the default system compiler (which seems to be gcc for me):

llama.cpp/Makefile

Lines 404 to 424 in c091cdf

ifdef LLAMA_HIPBLAS
ROCM_PATH ?= /opt/rocm
HIPCC ?= $(ROCM_PATH)/bin/hipcc
GPU_TARGETS ?= $(shell $(ROCM_PATH)/llvm/bin/amdgpu-arch)
LLAMA_CUDA_DMMV_X ?= 32
LLAMA_CUDA_MMV_Y ?= 1
LLAMA_CUDA_KQUANTS_ITER ?= 2
MK_CPPFLAGS += -DGGML_USE_HIPBLAS -DGGML_USE_CUBLAS
MK_LDFLAGS += -L$(ROCM_PATH)/lib -Wl,-rpath=$(ROCM_PATH)/lib
MK_LDFLAGS += -lhipblas -lamdhip64 -lrocblas
HIPFLAGS += $(addprefix --offload-arch=,$(GPU_TARGETS))
HIPFLAGS += -DGGML_CUDA_DMMV_X=$(LLAMA_CUDA_DMMV_X)
HIPFLAGS += -DGGML_CUDA_MMV_Y=$(LLAMA_CUDA_MMV_Y)
HIPFLAGS += -DK_QUANTS_PER_ITERATION=$(LLAMA_CUDA_KQUANTS_ITER)
ifdef LLAMA_CUDA_FORCE_DMMV
HIPFLAGS += -DGGML_CUDA_FORCE_DMMV
endif # LLAMA_CUDA_FORCE_DMMV
OBJS += ggml-cuda.o
ggml-cuda.o: ggml-cuda.cu ggml-cuda.h
$(HIPCC) $(CXXFLAGS) $(HIPFLAGS) -x hip -c -o $@ $<
endif # LLAMA_HIPBLAS

However it does use $(ROCM_PATH)/bin/hipcc instead of $(ROCM_PATH)/llvm/bin/clang / $(ROCM_PATH)/llvm/bin/clang++. Not sure what the difference is though. Unfortunately I also don't really know how to replicate that behavior with cmake. When I try only forcing CXX to the same compiler used in the makefile for ggml-cuda (CXX=/opt/rocm/bin/hipcc cmake .. -DLLAMA_HIPBLAS=ON), the build does go through and executing main now reports main: built with cc (GCC) 13.2.1 20230801 for x86_64-pc-linux-gnu, however it still ends with the same CUDA error 98.

@Displacer
Copy link
Author

I can confirm make LLAMA_HIPBLAS=1 and inference working without problem.
In case of make build ggml-cuda.cu is built with /opt/rocm/bin/hipcc
in case of cmake build ggml-cuda.cu is built with /opt/rocm/llvm/bin/clang++
It means there should be a rule to make ggml-cuda.cu with /opt/rocm/bin/hipcc in CMakeLists.txt

Thanks staviq for your support

@Displacer
Copy link
Author

Unfortunately some strange issue still exists:
look at the end of attached log file:

Instruction: write hello world in c \n### Assistant:Here is a simple C program that prints "Hello World":

#include <stdio.h>
int main() {
   printf("Hello, World!") ;
}

This code includes the standard input and output library (stdio.h) which contains the basic building basic syntax for the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the the 1999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999
output.txt

@staviq
Copy link
Contributor

staviq commented Sep 24, 2023

You should have main.xxx.log file, generated from that run, post it if you can

@Displacer
Copy link
Author

You should have main.xxx.log file, generated from that run, post it if you can

main.139793997638656.log.gz

@staviq
Copy link
Contributor

staviq commented Sep 24, 2023

Ok, that word puke at the end does come from the model, can you try with this one ?

https://huggingface.co/SlyEcho/open_llama_3b_v2_gguf/tree/main

@Displacer
Copy link
Author

Displacer commented Sep 24, 2023 •
edited
Loading

Ok, that word puke at the end does come from the model,

Not sure output from the phind-codellama-34b-v2.Q4_K_M.gguf model is wrong since it works perfect on pure cpu.

can you try with this one ?

https://huggingface.co/SlyEcho/open_llama_3b_v2_gguf/tree/main

This model generates simple hello world in c, but with more complicated query generated 265Mb log:

main.140070989163520.log.bz2.gz

@Displacer
Copy link
Author

Tested some queries inference with rocm 5.6 in new docker container instead of 5.7.

phind-codellama-34b-v2.Q4_K_M.gguf seems working without errors on GPU. open-llama-3b-v2-q5_1.gguf seems fine also.

Remaining llama.cpp issue is to patch CMakeLists.txt for using hipcc instead of clang++ for ggml-cuda.cu as it is done in simple make

Note rocm5.7 have some bug making models output invalid in most complicated cases.

@Engininja2
Copy link
Contributor

The makefile calls $(shell $(ROCM_PATH)/llvm/bin/amdgpu-arch) to find out what gpu archs are present on your PC to build for if you don't specify them manually. cmake leaves it up to the defaults in the hip-config.cmake file, which don't include gfx1100. If you configure cmake with -DAMDGPU_TARGETS=gfx1100 or the equivalent for your gpu (assuming hipblas supports it) it should work.

@Dasug
Copy link

Dasug commented Sep 26, 2023

Can confirm that for me at least the cuda error 98 error no longer occurs for me when building with cmake and -DAMDGPU_TARGETS=gfx1100.

@Dasug Dasug mentioned this issue Sep 26, 2023
Open
@the-snowwhite the-snowwhite mentioned this issue Oct 7, 2023
Open
@deadmeu deadmeu mentioned this issue Oct 9, 2023
Closed
@hugo-brites hugo-brites mentioned this issue Oct 9, 2023
Open
@Tungsten842 Tungsten842 mentioned this issue Oct 29, 2023
Merged
@ekg
Copy link
Contributor

ekg commented Dec 13, 2023

I have a similar experience with gfx1035. I can compile like this:

CC=/opt/rocm/llvm/bin/clang CXX=/opt/rocm/llvm/bin/clang++ cmake -H. -Bbuild -DLLAMA_HIPBLAS=ON -DAMDGPU_TARGETS=gfx1030 -DCMAKE_BUILD_TYPE=Release && cmake --build build -- -j 16

And run like this:

HSA_OVERRIDE_GFX_VERSION=10.3.0 build/bin/main -m ~/.cache/lm-studio/models/TheBloke/go-bruins-v2-GGUF/go-bruins-v2.Q5_K_M.gguf -p "good day"

And it runs, apparently using the GPU according to rocm-smi.

Now if I can just figure out how to offload more memory and compute to the GPU. It seems to use the VRAM for a scratch space? I can't set -ngl 1, I'll get an OOM for GPU.

@Alicimo Alicimo mentioned this issue Mar 2, 2024
Closed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging a pull request may close this issue.

flake.nix: fix for rocm 5.7 Tungsten842/llama.cpp
5 participants
@ekg @Displacer @Dasug @staviq @Engininja2