Image-Text-to-Text
Safetensors
qwen3_5
qwen3.8
quantized
gptq-int4-g32
conversational
4-bit precision
gptq
Instructions to use Chungulus/Qwen3.8-27B-GPTQ-INT4-G32 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Local Apps Settings
- vLLM
How to use Chungulus/Qwen3.8-27B-GPTQ-INT4-G32 with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "Chungulus/Qwen3.8-27B-GPTQ-INT4-G32" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "Chungulus/Qwen3.8-27B-GPTQ-INT4-G32", "messages": [ { "role": "user", "content": [ { "type": "text", "text": "Describe this image in one sentence." }, { "type": "image_url", "image_url": { "url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg" } } ] } ] }'Use Docker
docker model run hf.co/Chungulus/Qwen3.8-27B-GPTQ-INT4-G32
- SGLang
How to use Chungulus/Qwen3.8-27B-GPTQ-INT4-G32 with SGLang:
Install from pip and serve model
# Install SGLang from pip: pip install sglang # Start the SGLang server: python3 -m sglang.launch_server \ --model-path "Chungulus/Qwen3.8-27B-GPTQ-INT4-G32" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "Chungulus/Qwen3.8-27B-GPTQ-INT4-G32", "messages": [ { "role": "user", "content": [ { "type": "text", "text": "Describe this image in one sentence." }, { "type": "image_url", "image_url": { "url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg" } } ] } ] }'Use Docker images
docker run --gpus all \ --shm-size 32g \ -p 30000:30000 \ -v ~/.cache/huggingface:/root/.cache/huggingface \ --env "HF_TOKEN=<secret>" \ --ipc=host \ lmsysorg/sglang:latest \ python3 -m sglang.launch_server \ --model-path "Chungulus/Qwen3.8-27B-GPTQ-INT4-G32" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "Chungulus/Qwen3.8-27B-GPTQ-INT4-G32", "messages": [ { "role": "user", "content": [ { "type": "text", "text": "Describe this image in one sentence." }, { "type": "image_url", "image_url": { "url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg" } } ] } ] }' - Docker Model Runner
How to use Chungulus/Qwen3.8-27B-GPTQ-INT4-G32 with Docker Model Runner:
docker model run hf.co/Chungulus/Qwen3.8-27B-GPTQ-INT4-G32
Download quant_log.csv from Chungulus/Qwen3.8-27B-GPTQ-INT4-G32: direct link, hf CLI and curl.
- Browser
- Download file 19.4 kB
-
https://huggingface.co/Chungulus/Qwen3.8-27B-GPTQ-INT4-G32/resolve/main/quant_log.csv
- Command line
-
hf download hf://Chungulus/Qwen3.8-27B-GPTQ-INT4-G32/quant_log.csv
-
curl -L -o quant_log.csv https://huggingface.co/Chungulus/Qwen3.8-27B-GPTQ-INT4-G32/resolve/main/quant_log.csv
19.4 kB
| layer,module,loss,samples,damp,time | |
| 0,linear_attn.in_proj_qkv,0.0000465595,0.01000,4.838 | |
| 0,linear_attn.in_proj_z,0.0000305803,0.01000,3.571 | |
| 0,linear_attn.out_proj,0.0000000691,0.01000,3.946 | |
| 0,mlp.up_proj,0.0000002992,0.01000,9.341 | |
| 0,mlp.gate_proj,0.0000003201,0.01000,9.540 | |
| 0,mlp.down_proj,0.0000000011,0.01000,13.401 | |
| 1,linear_attn.in_proj_qkv,0.0000048377,0.01000,3.311 | |
| 1,linear_attn.in_proj_z,0.0000029758,0.01000,3.544 | |
| 1,linear_attn.out_proj,0.0000000068,0.01000,4.353 | |
| 1,mlp.up_proj,0.0000006274,0.01000,9.839 | |
| 1,mlp.gate_proj,0.0000006530,0.01000,9.897 | |
| 1,mlp.down_proj,0.0000000016,0.01000,13.298 | |
| 2,linear_attn.in_proj_qkv,0.0000065660,0.01000,3.461 | |
| 2,linear_attn.in_proj_z,0.0000039966,0.01000,3.271 | |
| 2,linear_attn.out_proj,0.0000000109,0.01000,4.008 | |
| 2,mlp.up_proj,0.0000011682,0.01000,10.524 | |
| 2,mlp.gate_proj,0.0000012390,0.01000,10.542 | |
| 2,mlp.down_proj,0.0000000053,0.01000,12.445 | |
| 3,self_attn.k_proj,0.0000071029,0.01000,17.450 | |
| 3,self_attn.q_proj,0.0001030882,0.01000,17.632 | |
| 3,self_attn.v_proj,0.0000068607,0.01000,17.771 | |
| 3,self_attn.o_proj,0.0000000149,0.01000,3.971 | |
| 3,mlp.up_proj,0.0000013900,0.01000,9.998 | |
| 3,mlp.gate_proj,0.0000014284,0.01000,10.084 | |
| 3,mlp.down_proj,0.0000000040,0.01000,13.235 | |
| 4,linear_attn.in_proj_qkv,0.0000068506,0.01000,4.197 | |
| 4,linear_attn.in_proj_z,0.0000043613,0.01000,3.714 | |
| 4,linear_attn.out_proj,0.0000000150,0.01000,3.891 | |
| 4,mlp.gate_proj,0.0000019025,0.01000,8.847 | |
| 4,mlp.up_proj,0.0000018767,0.01000,8.865 | |
| 4,mlp.down_proj,0.0000000052,0.01000,14.583 | |
| 5,linear_attn.in_proj_qkv,0.0000096762,0.01000,3.360 | |
| 5,linear_attn.in_proj_z,0.0000063109,0.01000,3.823 | |
| 5,linear_attn.out_proj,0.0000000166,0.01000,4.406 | |
| 5,mlp.up_proj,0.0000023280,0.01000,9.872 | |
| 5,mlp.gate_proj,0.0000024457,0.01000,9.909 | |
| 5,mlp.down_proj,0.0000000076,0.01000,12.991 | |
| 6,linear_attn.in_proj_qkv,0.0000172638,0.01000,3.553 | |
| 6,linear_attn.in_proj_z,0.0000103571,0.01000,3.410 | |
| 6,linear_attn.out_proj,0.0000000315,0.01000,3.838 | |
| 6,mlp.up_proj,0.0000034171,0.01000,10.269 | |
| 6,mlp.gate_proj,0.0000036210,0.01000,10.413 | |
| 6,mlp.down_proj,0.0000000134,0.01000,13.372 | |
| 7,self_attn.k_proj,0.0000059113,0.01000,17.050 | |
| 7,self_attn.q_proj,0.0000787790,0.01000,17.114 | |
| 7,self_attn.v_proj,0.0000055630,0.01000,17.235 | |
| 7,self_attn.o_proj,0.0000000369,0.01000,3.965 | |
| 7,mlp.up_proj,0.0000038006,0.01000,9.871 | |
| 7,mlp.gate_proj,0.0000040251,0.01000,9.951 | |
| 7,mlp.down_proj,0.0000000171,0.01000,13.071 | |
| 8,linear_attn.in_proj_qkv,0.0000140466,0.01000,3.524 | |
| 8,linear_attn.in_proj_z,0.0000088766,0.01000,3.591 | |
| 8,linear_attn.out_proj,0.0000000424,0.01000,3.995 | |
| 8,mlp.up_proj,0.0000042088,0.01000,8.027 | |
| 8,mlp.gate_proj,0.0000044353,0.01000,8.224 | |
| 8,mlp.down_proj,0.0000000186,0.01000,12.656 | |
| 9,linear_attn.in_proj_qkv,0.0000146743,0.01000,3.415 | |
| 9,linear_attn.in_proj_z,0.0000090112,0.01000,3.189 | |
| 9,linear_attn.out_proj,0.0000000407,0.01000,4.258 | |
| 9,mlp.up_proj,0.0000044021,0.01000,9.325 | |
| 9,mlp.gate_proj,0.0000046012,0.01000,9.398 | |
| 9,mlp.down_proj,0.0000000203,0.01000,11.828 | |
| 10,linear_attn.in_proj_qkv,0.0000148848,0.01000,3.323 | |
| 10,linear_attn.in_proj_z,0.0000089107,0.01000,3.390 | |
| 10,linear_attn.out_proj,0.0000000605,0.01000,4.166 | |
| 10,mlp.up_proj,0.0000044844,0.01000,9.541 | |
| 10,mlp.gate_proj,0.0000046451,0.01000,9.599 | |
| 10,mlp.down_proj,0.0000000244,0.01000,12.810 | |
| 11,self_attn.k_proj,0.0000049339,0.01000,16.118 | |
| 11,self_attn.q_proj,0.0000630472,0.01000,16.206 | |
| 11,self_attn.v_proj,0.0000047324,0.01000,16.319 | |
| 11,self_attn.o_proj,0.0000001008,0.01000,4.058 | |
| 11,mlp.up_proj,0.0000052183,0.01000,8.230 | |
| 11,mlp.gate_proj,0.0000054071,0.01000,8.277 | |
| 11,mlp.down_proj,0.0000000347,0.01000,12.730 | |
| 12,linear_attn.in_proj_qkv,0.0000193344,0.01000,4.267 | |
| 12,linear_attn.in_proj_z,0.0000115214,0.01000,4.999 | |
| 12,linear_attn.out_proj,0.0000000890,0.01000,5.844 | |
| 12,mlp.up_proj,0.0000054587,0.01000,10.800 | |
| 12,mlp.gate_proj,0.0000056580,0.01000,10.822 | |
| 12,mlp.down_proj,0.0000000387,0.01000,17.711 | |
| 13,linear_attn.in_proj_qkv,0.0000205298,0.01000,4.015 | |
| 13,linear_attn.in_proj_z,0.0000116700,0.01000,4.442 | |
| 13,linear_attn.out_proj,0.0000001050,0.01000,5.942 | |
| 13,mlp.up_proj,0.0000060140,0.01000,11.864 | |
| 13,mlp.gate_proj,0.0000062457,0.01000,11.944 | |
| 13,mlp.down_proj,0.0000000448,0.01000,17.853 | |
| 14,linear_attn.in_proj_qkv,0.0000228385,0.01000,3.808 | |
| 14,linear_attn.in_proj_z,0.0000134632,0.01000,4.232 | |
| 14,linear_attn.out_proj,0.0000001706,0.01000,6.292 | |
| 14,mlp.gate_proj,0.0000068234,0.01000,11.267 | |
| 14,mlp.up_proj,0.0000065948,0.01000,11.368 | |
| 14,mlp.down_proj,0.0000000568,0.01000,17.929 | |
| 15,self_attn.k_proj,0.0000042516,0.01000,20.526 | |
| 15,self_attn.v_proj,0.0000042227,0.01000,20.863 | |
| 15,self_attn.q_proj,0.0000498335,0.01000,20.935 | |
| 15,self_attn.o_proj,0.0000001376,0.01000,5.051 | |
| 15,mlp.up_proj,0.0000080230,0.01000,11.382 | |
| 15,mlp.gate_proj,0.0000082196,0.01000,11.410 | |
| 15,mlp.down_proj,0.0000000800,0.01000,16.850 | |
| 16,linear_attn.in_proj_qkv,0.0000244958,0.01000,4.882 | |
| 16,linear_attn.in_proj_z,0.0000139996,0.01000,4.429 | |
| 16,linear_attn.out_proj,0.0000001649,0.01000,4.950 | |
| 16,mlp.up_proj,0.0000090578,0.01000,11.231 | |
| 16,mlp.gate_proj,0.0000093029,0.01000,11.275 | |
| 16,mlp.down_proj,0.0000001030,0.01000,16.792 | |
| 17,linear_attn.in_proj_qkv,0.0000282585,0.01000,5.239 | |
| 17,linear_attn.in_proj_z,0.0000146324,0.01000,4.718 | |
| 17,linear_attn.out_proj,0.0000002194,0.01000,5.066 | |
| 17,mlp.gate_proj,0.0000110414,0.01000,11.310 | |
| 17,mlp.up_proj,0.0000107577,0.01000,11.355 | |
| 17,mlp.down_proj,0.0000001347,0.01000,16.906 | |
| 18,linear_attn.in_proj_qkv,0.0000304728,0.01000,4.571 | |
| 18,linear_attn.in_proj_z,0.0000170016,0.01000,4.615 | |
| 18,linear_attn.out_proj,0.0000003511,0.01000,5.093 | |
| 18,mlp.up_proj,0.0000138877,0.01000,11.130 | |
| 18,mlp.gate_proj,0.0000143212,0.01000,11.158 | |
| 18,mlp.down_proj,0.0000002540,0.01000,17.289 | |
| 19,self_attn.k_proj,0.0000044707,0.01000,19.904 | |
| 19,self_attn.q_proj,0.0000538362,0.01000,20.042 | |
| 19,self_attn.v_proj,0.0000048465,0.01000,20.141 | |
| 19,self_attn.o_proj,0.0000008460,0.01000,4.935 | |
| 19,mlp.gate_proj,0.0000218919,0.01000,12.212 | |
| 19,mlp.up_proj,0.0000211867,0.01000,12.329 | |
| 19,mlp.down_proj,0.0000004457,0.01000,16.790 | |
| 20,linear_attn.in_proj_qkv,0.0000626688,0.01000,3.637 | |
| 20,linear_attn.in_proj_z,0.0000359357,0.01000,4.110 | |
| 20,linear_attn.out_proj,0.0000007770,0.01000,5.139 | |
| 20,mlp.gate_proj,0.0000283249,0.01000,12.657 | |
| 20,mlp.up_proj,0.0000268738,0.01000,12.850 | |
| 20,mlp.down_proj,0.0000006253,0.01000,16.270 | |
| 21,linear_attn.in_proj_qkv,0.0000755664,0.01000,4.286 | |
| 21,linear_attn.in_proj_z,0.0000491567,0.01000,3.860 | |
| 21,linear_attn.out_proj,0.0000008459,0.01000,5.031 | |
| 21,mlp.up_proj,0.0000305040,0.01000,12.990 | |
| 21,mlp.gate_proj,0.0000320680,0.01000,13.029 | |
| 21,mlp.down_proj,0.0000005691,0.01000,15.314 | |
| 22,linear_attn.in_proj_qkv,0.0000670322,0.01000,4.261 | |
| 22,linear_attn.in_proj_z,0.0000417217,0.01000,4.178 | |
| 22,linear_attn.out_proj,0.0000006635,0.01000,5.132 | |
| 22,mlp.gate_proj,0.0000388471,0.01000,12.702 | |
| 22,mlp.up_proj,0.0000351618,0.01000,12.860 | |
| 22,mlp.down_proj,0.0000006516,0.01000,15.353 | |
| 23,self_attn.k_proj,0.0000069196,0.01000,19.722 | |
| 23,self_attn.v_proj,0.0000077916,0.01000,19.821 | |
| 23,self_attn.q_proj,0.0000793075,0.01000,19.915 | |
| 23,self_attn.o_proj,0.0000011365,0.01000,5.853 | |
| 23,mlp.up_proj,0.0000435874,0.01000,12.105 | |
| 23,mlp.gate_proj,0.0000463842,0.01000,12.134 | |
| 23,mlp.down_proj,0.0000008425,0.01000,16.878 | |
| 24,linear_attn.in_proj_qkv,0.0000698582,0.01000,5.795 | |
| 24,linear_attn.in_proj_z,0.0000404147,0.01000,3.633 | |
| 24,linear_attn.out_proj,0.0000012276,0.01000,5.637 | |
| 24,mlp.gate_proj,0.0000538200,0.01000,12.149 | |
| 24,mlp.up_proj,0.0000504490,0.01000,12.406 | |
| 24,mlp.down_proj,0.0000010879,0.01000,17.926 | |
| 25,linear_attn.in_proj_qkv,0.0000758185,0.01000,4.323 | |
| 25,linear_attn.in_proj_z,0.0000412498,0.01000,4.245 | |
| 25,linear_attn.out_proj,0.0000013347,0.01000,5.787 | |
| 25,mlp.gate_proj,0.0000522937,0.01000,11.877 | |
| 25,mlp.up_proj,0.0000513961,0.01000,11.993 | |
| 25,mlp.down_proj,0.0000012097,0.01000,17.317 | |
| 26,linear_attn.in_proj_qkv,0.0000698884,0.01000,4.032 | |
| 26,linear_attn.in_proj_z,0.0000367971,0.01000,4.119 | |
| 26,linear_attn.out_proj,0.0000021574,0.01000,5.719 | |
| 26,mlp.gate_proj,0.0000521411,0.01000,11.782 | |
| 26,mlp.up_proj,0.0000537349,0.01000,11.866 | |
| 26,mlp.down_proj,0.0000015733,0.01000,17.857 | |
| 27,self_attn.k_proj,0.0000077329,0.01000,20.770 | |
| 27,self_attn.q_proj,0.0000883736,0.01000,21.046 | |
| 27,self_attn.v_proj,0.0000080222,0.01000,21.144 | |
| 27,self_attn.o_proj,0.0000039121,0.01000,4.788 | |
| 27,mlp.up_proj,0.0000675765,0.01000,11.633 | |
| 27,mlp.gate_proj,0.0000635203,0.01000,11.692 | |
| 27,mlp.down_proj,0.0000022758,0.01000,17.126 | |
| 28,linear_attn.in_proj_qkv,0.0000878155,0.01000,4.772 | |
| 28,linear_attn.in_proj_z,0.0000424791,0.01000,4.331 | |
| 28,linear_attn.out_proj,0.0000027610,0.01000,5.341 | |
| 28,mlp.gate_proj,0.0000662915,0.01000,11.470 | |
| 28,mlp.up_proj,0.0000719431,0.01000,11.574 | |
| 28,mlp.down_proj,0.0000024675,0.01000,17.238 | |
| 29,linear_attn.in_proj_qkv,0.0001030115,0.01000,4.945 | |
| 29,linear_attn.in_proj_z,0.0000482778,0.01000,4.953 | |
| 29,linear_attn.out_proj,0.0000029918,0.01000,5.085 | |
| 29,mlp.gate_proj,0.0000713329,0.01000,11.608 | |
| 29,mlp.up_proj,0.0000777051,0.01000,11.647 | |
| 29,mlp.down_proj,0.0000029918,0.01000,17.246 | |
| 30,linear_attn.in_proj_qkv,0.0001099097,0.01000,5.325 | |
| 30,linear_attn.in_proj_z,0.0000528073,0.01000,4.685 | |
| 30,linear_attn.out_proj,0.0000046937,0.01000,4.467 | |
| 30,mlp.gate_proj,0.0000766497,0.01000,11.782 | |
| 30,mlp.up_proj,0.0000854201,0.01000,11.896 | |
| 30,mlp.down_proj,0.0000040538,0.01000,17.052 | |
| 31,self_attn.k_proj,0.0000117875,0.01000,20.046 | |
| 31,self_attn.v_proj,0.0000146597,0.01000,20.150 | |
| 31,self_attn.q_proj,0.0001238897,0.01000,20.185 | |
| 31,self_attn.o_proj,0.0000079213,0.01000,5.054 | |
| 31,mlp.gate_proj,0.0000901246,0.01000,12.966 | |
| 31,mlp.up_proj,0.0001026148,0.01000,12.995 | |
| 31,mlp.down_proj,0.0000054125,0.01000,15.526 | |
| 32,linear_attn.in_proj_qkv,0.0001411392,0.01000,4.297 | |
| 32,linear_attn.in_proj_z,0.0000633579,0.01000,4.290 | |
| 32,linear_attn.out_proj,0.0000059818,0.01000,4.378 | |
| 32,mlp.gate_proj,0.0001081419,0.01000,13.008 | |
| 32,mlp.up_proj,0.0001243875,0.01000,13.039 | |
| 32,mlp.down_proj,0.0000065014,0.01000,16.164 | |
| 33,linear_attn.in_proj_qkv,0.0001823122,0.01000,4.205 | |
| 33,linear_attn.in_proj_z,0.0000766750,0.01000,4.380 | |
| 33,linear_attn.out_proj,0.0000079385,0.01000,4.661 | |
| 33,mlp.gate_proj,0.0001244953,0.01000,12.917 | |
| 33,mlp.up_proj,0.0001456266,0.01000,12.994 | |
| 33,mlp.down_proj,0.0000071197,0.01000,16.185 | |
| 34,linear_attn.in_proj_qkv,0.0001931272,0.01000,4.665 | |
| 34,linear_attn.in_proj_z,0.0000884646,0.01000,4.202 | |
| 34,linear_attn.out_proj,0.0000076319,0.01000,5.294 | |
| 34,mlp.gate_proj,0.0001349536,0.01000,11.966 | |
| 34,mlp.up_proj,0.0001622027,0.01000,12.144 | |
| 34,mlp.down_proj,0.0000092788,0.01000,17.727 | |
| 35,self_attn.v_proj,0.0000297213,0.01000,20.105 | |
| 35,self_attn.k_proj,0.0000202671,0.01000,20.288 | |
| 35,self_attn.q_proj,0.0002131211,0.01000,20.312 | |
| 35,self_attn.o_proj,0.0000165856,0.01000,6.004 | |
| 35,mlp.gate_proj,0.0001520905,0.01000,11.776 | |
| 35,mlp.up_proj,0.0001754468,0.01000,11.835 | |
| 35,mlp.down_proj,0.0000088191,0.01000,17.986 | |
| 36,linear_attn.in_proj_qkv,0.0002695622,0.01000,4.256 | |
| 36,linear_attn.in_proj_z,0.0001382797,0.01000,4.288 | |
| 36,linear_attn.out_proj,0.0000042044,0.01000,5.340 | |
| 36,mlp.gate_proj,0.0001673946,0.01000,11.956 | |
| 36,mlp.up_proj,0.0001709744,0.01000,12.124 | |
| 36,mlp.down_proj,0.0000069419,0.01000,19.004 | |
| 37,linear_attn.in_proj_qkv,0.0002804669,0.01000,4.277 | |
| 37,linear_attn.in_proj_z,0.0001635379,0.01000,5.277 | |
| 37,linear_attn.out_proj,0.0000043134,0.01000,5.981 | |
| 37,mlp.up_proj,0.0001598896,0.01000,11.962 | |
| 37,mlp.gate_proj,0.0001610927,0.01000,12.038 | |
| 37,mlp.down_proj,0.0000061502,0.01000,18.387 | |
| 38,linear_attn.in_proj_qkv,0.0002559712,0.01000,4.407 | |
| 38,linear_attn.in_proj_z,0.0001519346,0.01000,5.260 | |
| 38,linear_attn.out_proj,0.0000033474,0.01000,5.072 | |
| 38,mlp.gate_proj,0.0002102250,0.01000,11.566 | |
| 38,mlp.up_proj,0.0001819107,0.01000,11.609 | |
| 38,mlp.down_proj,0.0000064125,0.01000,17.788 | |
| 39,self_attn.v_proj,0.0000256012,0.01000,20.503 | |
| 39,self_attn.q_proj,0.0002164532,0.01000,20.587 | |
| 39,self_attn.k_proj,0.0000195338,0.01000,20.718 | |
| 39,self_attn.o_proj,0.0000100446,0.01000,5.098 | |
| 39,mlp.up_proj,0.0001702656,0.01000,12.953 | |
| 39,mlp.gate_proj,0.0001858761,0.01000,13.025 | |
| 39,mlp.down_proj,0.0000053651,0.01000,15.871 | |
| 40,linear_attn.in_proj_qkv,0.0002165669,0.01000,6.048 | |
| 40,linear_attn.in_proj_z,0.0001219398,0.01000,4.178 | |
| 40,linear_attn.out_proj,0.0000043043,0.01000,4.662 | |
| 40,mlp.gate_proj,0.0002016520,0.01000,12.306 | |
| 40,mlp.up_proj,0.0001809801,0.01000,12.418 | |
| 40,mlp.down_proj,0.0000061339,0.01000,16.490 | |
| 41,linear_attn.in_proj_qkv,0.0002043740,0.01000,5.530 | |
| 41,linear_attn.in_proj_z,0.0001144006,0.01000,4.321 | |
| 41,linear_attn.out_proj,0.0000034344,0.01000,4.390 | |
| 41,mlp.up_proj,0.0001675813,0.01000,12.233 | |
| 41,mlp.gate_proj,0.0001684914,0.01000,12.310 | |
| 41,mlp.down_proj,0.0000066101,0.01000,16.020 | |
| 42,linear_attn.in_proj_qkv,0.0001864982,0.01000,5.811 | |
| 42,linear_attn.in_proj_z,0.0001013298,0.01000,4.442 | |
| 42,linear_attn.out_proj,0.0000060871,0.01000,5.023 | |
| 42,mlp.gate_proj,0.0001540212,0.01000,11.552 | |
| 42,mlp.up_proj,0.0001637856,0.01000,11.681 | |
| 42,mlp.down_proj,0.0000082843,0.01000,17.642 | |
| 43,self_attn.k_proj,0.0000194965,0.01000,20.762 | |
| 43,self_attn.v_proj,0.0000267600,0.01000,20.900 | |
| 43,self_attn.q_proj,0.0002105088,0.01000,21.024 | |
| 43,self_attn.o_proj,0.0000190630,0.01000,4.308 | |
| 43,mlp.up_proj,0.0001904914,0.01000,13.082 | |
| 43,mlp.gate_proj,0.0001721011,0.01000,13.195 | |
| 43,mlp.down_proj,0.0000101658,0.01000,17.366 | |
| 44,linear_attn.in_proj_qkv,0.0002125827,0.01000,4.503 | |
| 44,linear_attn.in_proj_z,0.0001049343,0.01000,4.503 | |
| 44,linear_attn.out_proj,0.0000073438,0.01000,5.065 | |
| 44,mlp.gate_proj,0.0001767046,0.01000,12.428 | |
| 44,mlp.up_proj,0.0001980982,0.01000,12.561 | |
| 44,mlp.down_proj,0.0000107053,0.01000,16.416 | |
| 45,linear_attn.in_proj_qkv,0.0002301205,0.01000,4.100 | |
| 45,linear_attn.in_proj_z,0.0001133411,0.01000,4.269 | |
| 45,linear_attn.out_proj,0.0000067958,0.01000,5.488 | |
| 45,mlp.gate_proj,0.0001815128,0.01000,12.960 | |
| 45,mlp.up_proj,0.0002038248,0.01000,12.999 | |
| 45,mlp.down_proj,0.0000117689,0.01000,17.039 | |
| 46,linear_attn.in_proj_qkv,0.0002410166,0.01000,4.602 | |
| 46,linear_attn.in_proj_z,0.0001195199,0.01000,4.756 | |
| 46,linear_attn.out_proj,0.0000106593,0.01000,6.061 | |
| 46,mlp.gate_proj,0.0001838977,0.01000,13.260 | |
| 46,mlp.up_proj,0.0002108314,0.01000,13.330 | |
| 46,mlp.down_proj,0.0000135484,0.01000,19.642 | |
| 47,self_attn.k_proj,0.0000272281,0.01000,16.301 | |
| 47,self_attn.v_proj,0.0000459376,0.01000,16.311 | |
| 47,self_attn.q_proj,0.0002806311,0.01000,16.431 | |
| 47,self_attn.o_proj,0.0000264252,0.01000,4.321 | |
| 47,mlp.gate_proj,0.0002192224,0.01000,9.299 | |
| 47,mlp.up_proj,0.0002510076,0.01000,9.386 | |
| 47,mlp.down_proj,0.0000174404,0.01000,12.689 | |
| 48,linear_attn.in_proj_qkv,0.0002837483,0.01000,3.622 | |
| 48,linear_attn.in_proj_z,0.0001363739,0.01000,2.871 | |
| 48,linear_attn.out_proj,0.0000098116,0.01000,3.461 | |
| 48,mlp.up_proj,0.0002761787,0.01000,9.404 | |
| 48,mlp.gate_proj,0.0002389715,0.01000,9.490 | |
| 48,mlp.down_proj,0.0000191603,0.01000,11.944 | |
| 49,linear_attn.in_proj_qkv,0.0003382158,0.01000,3.781 | |
| 49,linear_attn.in_proj_z,0.0001466734,0.01000,2.962 | |
| 49,linear_attn.out_proj,0.0000118660,0.01000,3.948 | |
| 49,mlp.up_proj,0.0002787075,0.01000,8.931 | |
| 49,mlp.gate_proj,0.0002376087,0.01000,8.958 | |
| 49,mlp.down_proj,0.0000217703,0.01000,12.872 | |
| 50,linear_attn.in_proj_qkv,0.0003170064,0.01000,3.242 | |
| 50,linear_attn.in_proj_z,0.0001498102,0.01000,3.053 | |
| 50,linear_attn.out_proj,0.0000222148,0.01000,3.818 | |
| 50,mlp.gate_proj,0.0002705110,0.01000,9.625 | |
| 50,mlp.up_proj,0.0003089729,0.01000,9.685 | |
| 50,mlp.down_proj,0.0000379229,0.01000,12.234 | |
| 51,self_attn.q_proj,0.0003904938,0.01000,16.188 | |
| 51,self_attn.k_proj,0.0000367747,0.01000,16.288 | |
| 51,self_attn.v_proj,0.0000770984,0.01000,16.317 | |
| 51,self_attn.o_proj,0.0000675822,0.01000,3.936 | |
| 51,mlp.up_proj,0.0003886414,0.01000,9.197 | |
| 51,mlp.gate_proj,0.0003319729,0.01000,9.223 | |
| 51,mlp.down_proj,0.0000452783,0.01000,12.002 | |
| 52,linear_attn.in_proj_qkv,0.0004915262,0.01000,3.585 | |
| 52,linear_attn.in_proj_z,0.0002357129,0.01000,3.747 | |
| 52,linear_attn.out_proj,0.0000185822,0.01000,4.429 | |
| 52,mlp.gate_proj,0.0003765633,0.01000,7.676 | |
| 52,mlp.up_proj,0.0003840524,0.01000,7.829 | |
| 52,mlp.down_proj,0.0000358539,0.01000,12.909 | |
| 53,linear_attn.in_proj_qkv,0.0004868115,0.01000,2.499 | |
| 53,linear_attn.in_proj_z,0.0002424216,0.01000,3.313 | |
| 53,linear_attn.out_proj,0.0000161716,0.01000,3.860 | |
| 53,mlp.gate_proj,0.0004000879,0.01000,9.965 | |
| 53,mlp.up_proj,0.0003850227,0.01000,10.024 | |
| 53,mlp.down_proj,0.0000381373,0.01000,12.209 | |
| 54,linear_attn.in_proj_qkv,0.0004595161,0.01000,3.252 | |
| 54,linear_attn.in_proj_z,0.0002529994,0.01000,3.525 | |
| 54,linear_attn.out_proj,0.0000212727,0.01000,4.330 | |
| 54,mlp.up_proj,0.0004466511,0.01000,8.232 | |
| 54,mlp.gate_proj,0.0004936524,0.01000,8.393 | |
| 54,mlp.down_proj,0.0000530516,0.01000,13.039 | |
| 55,self_attn.q_proj,0.0004352777,0.01000,15.776 | |
| 55,self_attn.v_proj,0.0000914873,0.01000,15.829 | |
| 55,self_attn.k_proj,0.0000395639,0.01000,15.901 | |
| 55,self_attn.o_proj,0.0000424179,0.01000,4.714 | |
| 55,mlp.up_proj,0.0004478328,0.01000,9.086 | |
| 55,mlp.gate_proj,0.0004768694,0.01000,9.150 | |
| 55,mlp.down_proj,0.0000382846,0.01000,12.185 | |
| 56,linear_attn.in_proj_qkv,0.0004960684,0.01000,3.595 | |
| 56,linear_attn.in_proj_z,0.0002529142,0.01000,3.049 | |
| 56,linear_attn.out_proj,0.0000178074,0.01000,3.998 | |
| 56,mlp.gate_proj,0.0004939634,0.01000,10.351 | |
| 56,mlp.up_proj,0.0004486849,0.01000,10.394 | |
| 56,mlp.down_proj,0.0000280172,0.01000,12.913 | |
| 57,linear_attn.in_proj_qkv,0.0005098134,0.01000,3.675 | |
| 57,linear_attn.in_proj_z,0.0002639756,0.01000,6.484 | |
| 57,linear_attn.out_proj,0.0000143255,0.01000,4.131 | |
| 57,mlp.up_proj,0.0005055156,0.01000,8.871 | |
| 57,mlp.gate_proj,0.0005675868,0.01000,8.919 | |
| 57,mlp.down_proj,0.0000293966,0.01000,13.795 | |
| 58,linear_attn.in_proj_qkv,0.0004955187,0.01000,3.365 | |
| 58,linear_attn.in_proj_z,0.0002751105,0.01000,2.776 | |
| 58,linear_attn.out_proj,0.0000218265,0.01000,2.988 | |
| 58,mlp.up_proj,0.0006015578,0.01000,9.663 | |
| 58,mlp.gate_proj,0.0006751317,0.01000,9.741 | |
| 58,mlp.down_proj,0.0000488076,0.01000,12.374 | |
| 59,self_attn.k_proj,0.0000493040,0.01000,17.238 | |
| 59,self_attn.q_proj,0.0005161758,0.01000,17.711 | |
| 59,self_attn.v_proj,0.0001953404,0.01000,17.753 | |
| 59,self_attn.o_proj,0.0001022928,0.01000,4.176 | |
| 59,mlp.gate_proj,0.0006658686,0.01000,8.717 | |
| 59,mlp.up_proj,0.0006093653,0.01000,8.902 | |
| 59,mlp.down_proj,0.0000679011,0.01000,12.651 | |
| 60,linear_attn.in_proj_qkv,0.0005593422,0.01000,3.221 | |
| 60,linear_attn.in_proj_z,0.0002729036,0.01000,2.679 | |
| 60,linear_attn.out_proj,0.0000513208,0.01000,4.113 | |
| 60,mlp.up_proj,0.0005944245,0.01000,9.187 | |
| 60,mlp.gate_proj,0.0006348773,0.01000,9.241 | |
| 60,mlp.down_proj,0.0000745988,0.01000,12.303 | |
| 61,linear_attn.in_proj_qkv,0.0004403089,0.01000,3.552 | |
| 61,linear_attn.in_proj_z,0.0002546274,0.01000,3.195 | |
| 61,linear_attn.out_proj,0.0000778997,0.01000,4.322 | |
| 61,mlp.gate_proj,0.0006772048,0.01000,9.103 | |
| 61,mlp.up_proj,0.0006411800,0.01000,9.131 | |
| 61,mlp.down_proj,0.0001109535,0.01000,12.444 | |
| 62,linear_attn.in_proj_qkv,0.0004666405,0.01000,3.666 | |
| 62,linear_attn.in_proj_z,0.0002469899,0.01000,2.752 | |
| 62,linear_attn.out_proj,0.0002129779,0.01000,3.847 | |
| 62,mlp.gate_proj,0.0006328930,0.01000,8.672 | |
| 62,mlp.up_proj,0.0006014320,0.01000,8.837 | |
| 62,mlp.down_proj,0.0001837080,0.01000,12.573 | |
| 63,self_attn.q_proj,0.0005039408,0.01000,16.709 | |
| 63,self_attn.k_proj,0.0000442387,0.01000,16.856 | |
| 63,self_attn.v_proj,0.0001613512,0.01000,16.868 | |
| 63,self_attn.o_proj,0.0002461547,0.01000,4.187 | |
| 63,mlp.gate_proj,0.0005901742,0.01000,9.777 | |
| 63,mlp.up_proj,0.0005432940,0.01000,9.816 | |
| 63,mlp.down_proj,0.0004097096,0.01000,12.570 | |