Text Generation
Transformers
Safetensors
qwen2
deepseek
int8
vllm
llmcompressor
conversational
text-generation-inference
8-bit precision
compressed-tensors
Instructions to use RedHatAI/DeepSeek-R1-Distill-Qwen-14B-quantized.w8a8 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use RedHatAI/DeepSeek-R1-Distill-Qwen-14B-quantized.w8a8 with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="RedHatAI/DeepSeek-R1-Distill-Qwen-14B-quantized.w8a8") messages = [ {"role": "user", "content": "Who are you?"}, ] pipe(messages)# Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("RedHatAI/DeepSeek-R1-Distill-Qwen-14B-quantized.w8a8") model = AutoModelForCausalLM.from_pretrained("RedHatAI/DeepSeek-R1-Distill-Qwen-14B-quantized.w8a8", device_map="auto") messages = [ {"role": "user", "content": "Who are you?"}, ] inputs = tokenizer.apply_chat_template( messages, add_generation_prompt=True, tokenize=True, return_dict=True, return_tensors="pt", ).to(model.device) outputs = model.generate(**inputs, max_new_tokens=40) print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:])) - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use RedHatAI/DeepSeek-R1-Distill-Qwen-14B-quantized.w8a8 with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "RedHatAI/DeepSeek-R1-Distill-Qwen-14B-quantized.w8a8" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "RedHatAI/DeepSeek-R1-Distill-Qwen-14B-quantized.w8a8", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker
docker model run hf.co/RedHatAI/DeepSeek-R1-Distill-Qwen-14B-quantized.w8a8
- SGLang
How to use RedHatAI/DeepSeek-R1-Distill-Qwen-14B-quantized.w8a8 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 "RedHatAI/DeepSeek-R1-Distill-Qwen-14B-quantized.w8a8" \ --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": "RedHatAI/DeepSeek-R1-Distill-Qwen-14B-quantized.w8a8", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'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 "RedHatAI/DeepSeek-R1-Distill-Qwen-14B-quantized.w8a8" \ --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": "RedHatAI/DeepSeek-R1-Distill-Qwen-14B-quantized.w8a8", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }' - Docker Model Runner
How to use RedHatAI/DeepSeek-R1-Distill-Qwen-14B-quantized.w8a8 with Docker Model Runner:
docker model run hf.co/RedHatAI/DeepSeek-R1-Distill-Qwen-14B-quantized.w8a8
Remove sparsity_config fields from config.json
Browse files- config.json +2 -348
config.json
CHANGED
|
@@ -57,353 +57,7 @@
|
|
| 57 |
"kv_cache_scheme": null,
|
| 58 |
"quant_method": "compressed-tensors",
|
| 59 |
"quantization_status": "compressed",
|
| 60 |
-
"sparsity_config": {
|
| 61 |
-
"format": "dense",
|
| 62 |
-
"global_sparsity": 0.05093884343560666,
|
| 63 |
-
"ignore": [
|
| 64 |
-
"model.layers.0.self_attn.q_proj",
|
| 65 |
-
"model.layers.0.self_attn.k_proj",
|
| 66 |
-
"model.layers.0.self_attn.v_proj",
|
| 67 |
-
"model.layers.0.self_attn.o_proj",
|
| 68 |
-
"model.layers.0.mlp.gate_proj",
|
| 69 |
-
"model.layers.0.mlp.up_proj",
|
| 70 |
-
"model.layers.0.mlp.down_proj",
|
| 71 |
-
"model.layers.1.self_attn.o_proj",
|
| 72 |
-
"model.layers.2.self_attn.v_proj",
|
| 73 |
-
"model.layers.2.self_attn.o_proj",
|
| 74 |
-
"model.layers.2.mlp.up_proj",
|
| 75 |
-
"model.layers.3.self_attn.q_proj",
|
| 76 |
-
"model.layers.3.self_attn.k_proj",
|
| 77 |
-
"model.layers.3.self_attn.v_proj",
|
| 78 |
-
"model.layers.3.self_attn.o_proj",
|
| 79 |
-
"model.layers.3.mlp.up_proj",
|
| 80 |
-
"model.layers.3.mlp.down_proj",
|
| 81 |
-
"model.layers.4.self_attn.q_proj",
|
| 82 |
-
"model.layers.4.self_attn.k_proj",
|
| 83 |
-
"model.layers.4.self_attn.v_proj",
|
| 84 |
-
"model.layers.4.self_attn.o_proj",
|
| 85 |
-
"model.layers.4.mlp.gate_proj",
|
| 86 |
-
"model.layers.4.mlp.up_proj",
|
| 87 |
-
"model.layers.4.mlp.down_proj",
|
| 88 |
-
"model.layers.5.self_attn.q_proj",
|
| 89 |
-
"model.layers.5.self_attn.k_proj",
|
| 90 |
-
"model.layers.5.self_attn.v_proj",
|
| 91 |
-
"model.layers.5.self_attn.o_proj",
|
| 92 |
-
"model.layers.5.mlp.gate_proj",
|
| 93 |
-
"model.layers.5.mlp.up_proj",
|
| 94 |
-
"model.layers.5.mlp.down_proj",
|
| 95 |
-
"model.layers.6.self_attn.q_proj",
|
| 96 |
-
"model.layers.6.self_attn.k_proj",
|
| 97 |
-
"model.layers.6.self_attn.v_proj",
|
| 98 |
-
"model.layers.6.self_attn.o_proj",
|
| 99 |
-
"model.layers.6.mlp.gate_proj",
|
| 100 |
-
"model.layers.6.mlp.up_proj",
|
| 101 |
-
"model.layers.6.mlp.down_proj",
|
| 102 |
-
"model.layers.7.self_attn.q_proj",
|
| 103 |
-
"model.layers.7.self_attn.k_proj",
|
| 104 |
-
"model.layers.7.self_attn.v_proj",
|
| 105 |
-
"model.layers.7.self_attn.o_proj",
|
| 106 |
-
"model.layers.7.mlp.gate_proj",
|
| 107 |
-
"model.layers.7.mlp.up_proj",
|
| 108 |
-
"model.layers.7.mlp.down_proj",
|
| 109 |
-
"model.layers.8.self_attn.q_proj",
|
| 110 |
-
"model.layers.8.self_attn.k_proj",
|
| 111 |
-
"model.layers.8.self_attn.v_proj",
|
| 112 |
-
"model.layers.8.self_attn.o_proj",
|
| 113 |
-
"model.layers.8.mlp.gate_proj",
|
| 114 |
-
"model.layers.8.mlp.up_proj",
|
| 115 |
-
"model.layers.8.mlp.down_proj",
|
| 116 |
-
"model.layers.9.self_attn.q_proj",
|
| 117 |
-
"model.layers.9.self_attn.k_proj",
|
| 118 |
-
"model.layers.9.self_attn.v_proj",
|
| 119 |
-
"model.layers.9.self_attn.o_proj",
|
| 120 |
-
"model.layers.9.mlp.gate_proj",
|
| 121 |
-
"model.layers.9.mlp.up_proj",
|
| 122 |
-
"model.layers.9.mlp.down_proj",
|
| 123 |
-
"model.layers.10.self_attn.q_proj",
|
| 124 |
-
"model.layers.10.self_attn.k_proj",
|
| 125 |
-
"model.layers.10.self_attn.v_proj",
|
| 126 |
-
"model.layers.10.self_attn.o_proj",
|
| 127 |
-
"model.layers.10.mlp.gate_proj",
|
| 128 |
-
"model.layers.10.mlp.up_proj",
|
| 129 |
-
"model.layers.10.mlp.down_proj",
|
| 130 |
-
"model.layers.11.self_attn.q_proj",
|
| 131 |
-
"model.layers.11.self_attn.k_proj",
|
| 132 |
-
"model.layers.11.self_attn.v_proj",
|
| 133 |
-
"model.layers.11.self_attn.o_proj",
|
| 134 |
-
"model.layers.11.mlp.gate_proj",
|
| 135 |
-
"model.layers.11.mlp.up_proj",
|
| 136 |
-
"model.layers.11.mlp.down_proj",
|
| 137 |
-
"model.layers.12.self_attn.q_proj",
|
| 138 |
-
"model.layers.12.self_attn.k_proj",
|
| 139 |
-
"model.layers.12.self_attn.v_proj",
|
| 140 |
-
"model.layers.12.self_attn.o_proj",
|
| 141 |
-
"model.layers.12.mlp.gate_proj",
|
| 142 |
-
"model.layers.12.mlp.up_proj",
|
| 143 |
-
"model.layers.12.mlp.down_proj",
|
| 144 |
-
"model.layers.13.self_attn.q_proj",
|
| 145 |
-
"model.layers.13.self_attn.k_proj",
|
| 146 |
-
"model.layers.13.self_attn.v_proj",
|
| 147 |
-
"model.layers.13.self_attn.o_proj",
|
| 148 |
-
"model.layers.13.mlp.gate_proj",
|
| 149 |
-
"model.layers.13.mlp.up_proj",
|
| 150 |
-
"model.layers.13.mlp.down_proj",
|
| 151 |
-
"model.layers.14.self_attn.q_proj",
|
| 152 |
-
"model.layers.14.self_attn.k_proj",
|
| 153 |
-
"model.layers.14.self_attn.v_proj",
|
| 154 |
-
"model.layers.14.self_attn.o_proj",
|
| 155 |
-
"model.layers.14.mlp.gate_proj",
|
| 156 |
-
"model.layers.14.mlp.up_proj",
|
| 157 |
-
"model.layers.14.mlp.down_proj",
|
| 158 |
-
"model.layers.15.self_attn.q_proj",
|
| 159 |
-
"model.layers.15.self_attn.k_proj",
|
| 160 |
-
"model.layers.15.self_attn.v_proj",
|
| 161 |
-
"model.layers.15.self_attn.o_proj",
|
| 162 |
-
"model.layers.15.mlp.gate_proj",
|
| 163 |
-
"model.layers.15.mlp.up_proj",
|
| 164 |
-
"model.layers.15.mlp.down_proj",
|
| 165 |
-
"model.layers.16.self_attn.q_proj",
|
| 166 |
-
"model.layers.16.self_attn.k_proj",
|
| 167 |
-
"model.layers.16.self_attn.v_proj",
|
| 168 |
-
"model.layers.16.self_attn.o_proj",
|
| 169 |
-
"model.layers.16.mlp.gate_proj",
|
| 170 |
-
"model.layers.16.mlp.up_proj",
|
| 171 |
-
"model.layers.16.mlp.down_proj",
|
| 172 |
-
"model.layers.17.self_attn.q_proj",
|
| 173 |
-
"model.layers.17.self_attn.k_proj",
|
| 174 |
-
"model.layers.17.self_attn.v_proj",
|
| 175 |
-
"model.layers.17.self_attn.o_proj",
|
| 176 |
-
"model.layers.17.mlp.gate_proj",
|
| 177 |
-
"model.layers.17.mlp.up_proj",
|
| 178 |
-
"model.layers.17.mlp.down_proj",
|
| 179 |
-
"model.layers.18.self_attn.q_proj",
|
| 180 |
-
"model.layers.18.self_attn.k_proj",
|
| 181 |
-
"model.layers.18.self_attn.v_proj",
|
| 182 |
-
"model.layers.18.self_attn.o_proj",
|
| 183 |
-
"model.layers.18.mlp.gate_proj",
|
| 184 |
-
"model.layers.18.mlp.up_proj",
|
| 185 |
-
"model.layers.18.mlp.down_proj",
|
| 186 |
-
"model.layers.19.self_attn.q_proj",
|
| 187 |
-
"model.layers.19.self_attn.k_proj",
|
| 188 |
-
"model.layers.19.self_attn.v_proj",
|
| 189 |
-
"model.layers.19.self_attn.o_proj",
|
| 190 |
-
"model.layers.19.mlp.gate_proj",
|
| 191 |
-
"model.layers.19.mlp.up_proj",
|
| 192 |
-
"model.layers.19.mlp.down_proj",
|
| 193 |
-
"model.layers.20.self_attn.q_proj",
|
| 194 |
-
"model.layers.20.self_attn.k_proj",
|
| 195 |
-
"model.layers.20.self_attn.v_proj",
|
| 196 |
-
"model.layers.20.self_attn.o_proj",
|
| 197 |
-
"model.layers.20.mlp.gate_proj",
|
| 198 |
-
"model.layers.20.mlp.up_proj",
|
| 199 |
-
"model.layers.20.mlp.down_proj",
|
| 200 |
-
"model.layers.21.self_attn.q_proj",
|
| 201 |
-
"model.layers.21.self_attn.k_proj",
|
| 202 |
-
"model.layers.21.self_attn.v_proj",
|
| 203 |
-
"model.layers.21.self_attn.o_proj",
|
| 204 |
-
"model.layers.21.mlp.gate_proj",
|
| 205 |
-
"model.layers.21.mlp.up_proj",
|
| 206 |
-
"model.layers.21.mlp.down_proj",
|
| 207 |
-
"model.layers.22.self_attn.q_proj",
|
| 208 |
-
"model.layers.22.self_attn.k_proj",
|
| 209 |
-
"model.layers.22.self_attn.v_proj",
|
| 210 |
-
"model.layers.22.self_attn.o_proj",
|
| 211 |
-
"model.layers.22.mlp.gate_proj",
|
| 212 |
-
"model.layers.22.mlp.up_proj",
|
| 213 |
-
"model.layers.22.mlp.down_proj",
|
| 214 |
-
"model.layers.23.self_attn.q_proj",
|
| 215 |
-
"model.layers.23.self_attn.k_proj",
|
| 216 |
-
"model.layers.23.self_attn.v_proj",
|
| 217 |
-
"model.layers.23.self_attn.o_proj",
|
| 218 |
-
"model.layers.23.mlp.gate_proj",
|
| 219 |
-
"model.layers.23.mlp.up_proj",
|
| 220 |
-
"model.layers.23.mlp.down_proj",
|
| 221 |
-
"model.layers.24.self_attn.q_proj",
|
| 222 |
-
"model.layers.24.self_attn.k_proj",
|
| 223 |
-
"model.layers.24.self_attn.v_proj",
|
| 224 |
-
"model.layers.24.self_attn.o_proj",
|
| 225 |
-
"model.layers.24.mlp.gate_proj",
|
| 226 |
-
"model.layers.24.mlp.up_proj",
|
| 227 |
-
"model.layers.24.mlp.down_proj",
|
| 228 |
-
"model.layers.25.self_attn.q_proj",
|
| 229 |
-
"model.layers.25.self_attn.k_proj",
|
| 230 |
-
"model.layers.25.self_attn.v_proj",
|
| 231 |
-
"model.layers.25.self_attn.o_proj",
|
| 232 |
-
"model.layers.25.mlp.gate_proj",
|
| 233 |
-
"model.layers.25.mlp.up_proj",
|
| 234 |
-
"model.layers.25.mlp.down_proj",
|
| 235 |
-
"model.layers.26.self_attn.q_proj",
|
| 236 |
-
"model.layers.26.self_attn.k_proj",
|
| 237 |
-
"model.layers.26.self_attn.v_proj",
|
| 238 |
-
"model.layers.26.self_attn.o_proj",
|
| 239 |
-
"model.layers.26.mlp.gate_proj",
|
| 240 |
-
"model.layers.26.mlp.up_proj",
|
| 241 |
-
"model.layers.26.mlp.down_proj",
|
| 242 |
-
"model.layers.27.self_attn.q_proj",
|
| 243 |
-
"model.layers.27.self_attn.k_proj",
|
| 244 |
-
"model.layers.27.self_attn.v_proj",
|
| 245 |
-
"model.layers.27.self_attn.o_proj",
|
| 246 |
-
"model.layers.27.mlp.gate_proj",
|
| 247 |
-
"model.layers.27.mlp.up_proj",
|
| 248 |
-
"model.layers.27.mlp.down_proj",
|
| 249 |
-
"model.layers.28.self_attn.q_proj",
|
| 250 |
-
"model.layers.28.self_attn.k_proj",
|
| 251 |
-
"model.layers.28.self_attn.v_proj",
|
| 252 |
-
"model.layers.28.self_attn.o_proj",
|
| 253 |
-
"model.layers.28.mlp.gate_proj",
|
| 254 |
-
"model.layers.28.mlp.up_proj",
|
| 255 |
-
"model.layers.28.mlp.down_proj",
|
| 256 |
-
"model.layers.29.self_attn.q_proj",
|
| 257 |
-
"model.layers.29.self_attn.k_proj",
|
| 258 |
-
"model.layers.29.self_attn.v_proj",
|
| 259 |
-
"model.layers.29.self_attn.o_proj",
|
| 260 |
-
"model.layers.29.mlp.gate_proj",
|
| 261 |
-
"model.layers.29.mlp.up_proj",
|
| 262 |
-
"model.layers.29.mlp.down_proj",
|
| 263 |
-
"model.layers.30.self_attn.q_proj",
|
| 264 |
-
"model.layers.30.self_attn.k_proj",
|
| 265 |
-
"model.layers.30.self_attn.v_proj",
|
| 266 |
-
"model.layers.30.self_attn.o_proj",
|
| 267 |
-
"model.layers.30.mlp.gate_proj",
|
| 268 |
-
"model.layers.30.mlp.up_proj",
|
| 269 |
-
"model.layers.30.mlp.down_proj",
|
| 270 |
-
"model.layers.31.self_attn.q_proj",
|
| 271 |
-
"model.layers.31.self_attn.k_proj",
|
| 272 |
-
"model.layers.31.self_attn.v_proj",
|
| 273 |
-
"model.layers.31.self_attn.o_proj",
|
| 274 |
-
"model.layers.31.mlp.gate_proj",
|
| 275 |
-
"model.layers.31.mlp.up_proj",
|
| 276 |
-
"model.layers.31.mlp.down_proj",
|
| 277 |
-
"model.layers.32.self_attn.q_proj",
|
| 278 |
-
"model.layers.32.self_attn.k_proj",
|
| 279 |
-
"model.layers.32.self_attn.v_proj",
|
| 280 |
-
"model.layers.32.self_attn.o_proj",
|
| 281 |
-
"model.layers.32.mlp.gate_proj",
|
| 282 |
-
"model.layers.32.mlp.up_proj",
|
| 283 |
-
"model.layers.32.mlp.down_proj",
|
| 284 |
-
"model.layers.33.self_attn.q_proj",
|
| 285 |
-
"model.layers.33.self_attn.k_proj",
|
| 286 |
-
"model.layers.33.self_attn.v_proj",
|
| 287 |
-
"model.layers.33.self_attn.o_proj",
|
| 288 |
-
"model.layers.33.mlp.gate_proj",
|
| 289 |
-
"model.layers.33.mlp.up_proj",
|
| 290 |
-
"model.layers.33.mlp.down_proj",
|
| 291 |
-
"model.layers.34.self_attn.q_proj",
|
| 292 |
-
"model.layers.34.self_attn.k_proj",
|
| 293 |
-
"model.layers.34.self_attn.v_proj",
|
| 294 |
-
"model.layers.34.self_attn.o_proj",
|
| 295 |
-
"model.layers.34.mlp.gate_proj",
|
| 296 |
-
"model.layers.34.mlp.up_proj",
|
| 297 |
-
"model.layers.34.mlp.down_proj",
|
| 298 |
-
"model.layers.35.self_attn.q_proj",
|
| 299 |
-
"model.layers.35.self_attn.k_proj",
|
| 300 |
-
"model.layers.35.self_attn.v_proj",
|
| 301 |
-
"model.layers.35.self_attn.o_proj",
|
| 302 |
-
"model.layers.35.mlp.gate_proj",
|
| 303 |
-
"model.layers.35.mlp.up_proj",
|
| 304 |
-
"model.layers.35.mlp.down_proj",
|
| 305 |
-
"model.layers.36.self_attn.q_proj",
|
| 306 |
-
"model.layers.36.self_attn.k_proj",
|
| 307 |
-
"model.layers.36.self_attn.v_proj",
|
| 308 |
-
"model.layers.36.self_attn.o_proj",
|
| 309 |
-
"model.layers.36.mlp.gate_proj",
|
| 310 |
-
"model.layers.36.mlp.up_proj",
|
| 311 |
-
"model.layers.36.mlp.down_proj",
|
| 312 |
-
"model.layers.37.self_attn.q_proj",
|
| 313 |
-
"model.layers.37.self_attn.k_proj",
|
| 314 |
-
"model.layers.37.self_attn.v_proj",
|
| 315 |
-
"model.layers.37.self_attn.o_proj",
|
| 316 |
-
"model.layers.37.mlp.gate_proj",
|
| 317 |
-
"model.layers.37.mlp.up_proj",
|
| 318 |
-
"model.layers.37.mlp.down_proj",
|
| 319 |
-
"model.layers.38.self_attn.q_proj",
|
| 320 |
-
"model.layers.38.self_attn.k_proj",
|
| 321 |
-
"model.layers.38.self_attn.v_proj",
|
| 322 |
-
"model.layers.38.self_attn.o_proj",
|
| 323 |
-
"model.layers.38.mlp.gate_proj",
|
| 324 |
-
"model.layers.38.mlp.up_proj",
|
| 325 |
-
"model.layers.38.mlp.down_proj",
|
| 326 |
-
"model.layers.39.self_attn.q_proj",
|
| 327 |
-
"model.layers.39.self_attn.k_proj",
|
| 328 |
-
"model.layers.39.self_attn.v_proj",
|
| 329 |
-
"model.layers.39.self_attn.o_proj",
|
| 330 |
-
"model.layers.39.mlp.gate_proj",
|
| 331 |
-
"model.layers.39.mlp.up_proj",
|
| 332 |
-
"model.layers.39.mlp.down_proj",
|
| 333 |
-
"model.layers.40.self_attn.q_proj",
|
| 334 |
-
"model.layers.40.self_attn.k_proj",
|
| 335 |
-
"model.layers.40.self_attn.v_proj",
|
| 336 |
-
"model.layers.40.self_attn.o_proj",
|
| 337 |
-
"model.layers.40.mlp.gate_proj",
|
| 338 |
-
"model.layers.40.mlp.up_proj",
|
| 339 |
-
"model.layers.40.mlp.down_proj",
|
| 340 |
-
"model.layers.41.self_attn.q_proj",
|
| 341 |
-
"model.layers.41.self_attn.k_proj",
|
| 342 |
-
"model.layers.41.self_attn.v_proj",
|
| 343 |
-
"model.layers.41.self_attn.o_proj",
|
| 344 |
-
"model.layers.41.mlp.gate_proj",
|
| 345 |
-
"model.layers.41.mlp.up_proj",
|
| 346 |
-
"model.layers.41.mlp.down_proj",
|
| 347 |
-
"model.layers.42.self_attn.q_proj",
|
| 348 |
-
"model.layers.42.self_attn.k_proj",
|
| 349 |
-
"model.layers.42.self_attn.v_proj",
|
| 350 |
-
"model.layers.42.self_attn.o_proj",
|
| 351 |
-
"model.layers.42.mlp.gate_proj",
|
| 352 |
-
"model.layers.42.mlp.up_proj",
|
| 353 |
-
"model.layers.42.mlp.down_proj",
|
| 354 |
-
"model.layers.43.self_attn.q_proj",
|
| 355 |
-
"model.layers.43.self_attn.k_proj",
|
| 356 |
-
"model.layers.43.self_attn.v_proj",
|
| 357 |
-
"model.layers.43.self_attn.o_proj",
|
| 358 |
-
"model.layers.43.mlp.gate_proj",
|
| 359 |
-
"model.layers.43.mlp.up_proj",
|
| 360 |
-
"model.layers.43.mlp.down_proj",
|
| 361 |
-
"model.layers.44.self_attn.q_proj",
|
| 362 |
-
"model.layers.44.self_attn.k_proj",
|
| 363 |
-
"model.layers.44.self_attn.v_proj",
|
| 364 |
-
"model.layers.44.self_attn.o_proj",
|
| 365 |
-
"model.layers.44.mlp.gate_proj",
|
| 366 |
-
"model.layers.44.mlp.up_proj",
|
| 367 |
-
"model.layers.44.mlp.down_proj",
|
| 368 |
-
"model.layers.45.self_attn.q_proj",
|
| 369 |
-
"model.layers.45.self_attn.k_proj",
|
| 370 |
-
"model.layers.45.self_attn.v_proj",
|
| 371 |
-
"model.layers.45.self_attn.o_proj",
|
| 372 |
-
"model.layers.45.mlp.gate_proj",
|
| 373 |
-
"model.layers.45.mlp.up_proj",
|
| 374 |
-
"model.layers.45.mlp.down_proj",
|
| 375 |
-
"model.layers.46.self_attn.q_proj",
|
| 376 |
-
"model.layers.46.self_attn.k_proj",
|
| 377 |
-
"model.layers.46.self_attn.v_proj",
|
| 378 |
-
"model.layers.46.self_attn.o_proj",
|
| 379 |
-
"model.layers.46.mlp.gate_proj",
|
| 380 |
-
"model.layers.46.mlp.up_proj",
|
| 381 |
-
"model.layers.46.mlp.down_proj",
|
| 382 |
-
"model.layers.47.self_attn.q_proj",
|
| 383 |
-
"model.layers.47.self_attn.k_proj",
|
| 384 |
-
"model.layers.47.self_attn.v_proj",
|
| 385 |
-
"model.layers.47.self_attn.o_proj",
|
| 386 |
-
"model.layers.47.mlp.gate_proj",
|
| 387 |
-
"model.layers.47.mlp.up_proj",
|
| 388 |
-
"model.layers.47.mlp.down_proj",
|
| 389 |
-
"lm_head"
|
| 390 |
-
],
|
| 391 |
-
"registry_requires_subclass": false,
|
| 392 |
-
"sparsity_structure": "unstructured",
|
| 393 |
-
"targets": [
|
| 394 |
-
"model.layers.1.self_attn.q_proj",
|
| 395 |
-
"model.layers.1.self_attn.k_proj",
|
| 396 |
-
"model.layers.1.self_attn.v_proj",
|
| 397 |
-
"model.layers.1.mlp.gate_proj",
|
| 398 |
-
"model.layers.1.mlp.up_proj",
|
| 399 |
-
"model.layers.1.mlp.down_proj",
|
| 400 |
-
"model.layers.2.self_attn.q_proj",
|
| 401 |
-
"model.layers.2.self_attn.k_proj",
|
| 402 |
-
"model.layers.2.mlp.gate_proj",
|
| 403 |
-
"model.layers.2.mlp.down_proj",
|
| 404 |
-
"model.layers.3.mlp.gate_proj"
|
| 405 |
-
]
|
| 406 |
-
}
|
| 407 |
},
|
| 408 |
"rms_norm_eps": 1e-05,
|
| 409 |
"rope_scaling": null,
|
|
@@ -415,4 +69,4 @@
|
|
| 415 |
"use_cache": true,
|
| 416 |
"use_sliding_window": false,
|
| 417 |
"vocab_size": 152064
|
| 418 |
-
}
|
|
|
|
| 57 |
"kv_cache_scheme": null,
|
| 58 |
"quant_method": "compressed-tensors",
|
| 59 |
"quantization_status": "compressed",
|
| 60 |
+
"sparsity_config": {}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 61 |
},
|
| 62 |
"rms_norm_eps": 1e-05,
|
| 63 |
"rope_scaling": null,
|
|
|
|
| 69 |
"use_cache": true,
|
| 70 |
"use_sliding_window": false,
|
| 71 |
"vocab_size": 152064
|
| 72 |
+
}
|