Image-Text-to-Text
GGUF
gsq
rco
quantization
mixed-precision
veriloop
vision
multimodal
code
math
speculative-decoding
mtp
imatrix
conversational
Instructions to use tahaalam2009/VeriLoop-E2-GSQ-RCO-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- llama.cpp
How to use tahaalam2009/VeriLoop-E2-GSQ-RCO-GGUF with llama.cpp:
Install (macOS, Linux)
curl -LsSf https://llama.app/install.sh | sh # Start a local OpenAI-compatible server with a web UI: llama serve -hf tahaalam2009/VeriLoop-E2-GSQ-RCO-GGUF:IQ2_S # Run inference directly in the terminal: llama cli -hf tahaalam2009/VeriLoop-E2-GSQ-RCO-GGUF:IQ2_S
Install from WinGet (Windows)
winget install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama serve -hf tahaalam2009/VeriLoop-E2-GSQ-RCO-GGUF:IQ2_S # Run inference directly in the terminal: llama cli -hf tahaalam2009/VeriLoop-E2-GSQ-RCO-GGUF:IQ2_S
Use pre-built binary
# Download pre-built binary from: # https://github.com/ggerganov/llama.cpp/releases # Start a local OpenAI-compatible server with a web UI: ./llama-server -hf tahaalam2009/VeriLoop-E2-GSQ-RCO-GGUF:IQ2_S # Run inference directly in the terminal: ./llama-cli -hf tahaalam2009/VeriLoop-E2-GSQ-RCO-GGUF:IQ2_S
Build from source code
git clone https://github.com/ggerganov/llama.cpp.git cd llama.cpp cmake -B build cmake --build build -j --target llama-server llama-cli # Start a local OpenAI-compatible server with a web UI: ./build/bin/llama-server -hf tahaalam2009/VeriLoop-E2-GSQ-RCO-GGUF:IQ2_S # Run inference directly in the terminal: ./build/bin/llama-cli -hf tahaalam2009/VeriLoop-E2-GSQ-RCO-GGUF:IQ2_S
Use Docker
docker model run hf.co/tahaalam2009/VeriLoop-E2-GSQ-RCO-GGUF:IQ2_S
- LM Studio
- Jan
- vLLM
How to use tahaalam2009/VeriLoop-E2-GSQ-RCO-GGUF with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "tahaalam2009/VeriLoop-E2-GSQ-RCO-GGUF" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "tahaalam2009/VeriLoop-E2-GSQ-RCO-GGUF", "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/tahaalam2009/VeriLoop-E2-GSQ-RCO-GGUF:IQ2_S
- Ollama
How to use tahaalam2009/VeriLoop-E2-GSQ-RCO-GGUF with Ollama:
ollama run hf.co/tahaalam2009/VeriLoop-E2-GSQ-RCO-GGUF:IQ2_S
- Unsloth Desktop
- Pi
How to use tahaalam2009/VeriLoop-E2-GSQ-RCO-GGUF with Pi:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf tahaalam2009/VeriLoop-E2-GSQ-RCO-GGUF:IQ2_S
Configure the model in Pi
# Install Pi: npm install -g @earendil-works/pi-coding-agent # Add to ~/.pi/agent/models.json: { "providers": { "llama-cpp": { "baseUrl": "http://localhost:8080/v1", "api": "openai-completions", "apiKey": "none", "models": [ { "id": "tahaalam2009/VeriLoop-E2-GSQ-RCO-GGUF:IQ2_S" } ] } } }Run Pi
# Start Pi in your project directory: pi
- Docker Model Runner
How to use tahaalam2009/VeriLoop-E2-GSQ-RCO-GGUF with Docker Model Runner:
docker model run hf.co/tahaalam2009/VeriLoop-E2-GSQ-RCO-GGUF:IQ2_S
- Lemonade
How to use tahaalam2009/VeriLoop-E2-GSQ-RCO-GGUF with Lemonade:
Pull the model
# Download Lemonade from https://lemonade-server.ai/ lemonade pull tahaalam2009/VeriLoop-E2-GSQ-RCO-GGUF:IQ2_S
Run and chat with the model
lemonade run user.VeriLoop-E2-GSQ-RCO-GGUF-IQ2_S
List all available models
lemonade list
- Hermes Agent
How to use tahaalam2009/VeriLoop-E2-GSQ-RCO-GGUF with Hermes Agent:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf tahaalam2009/VeriLoop-E2-GSQ-RCO-GGUF:IQ2_S
Configure Hermes
# Install Hermes: curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash hermes setup # Point Hermes at the local server: hermes config set model.provider custom hermes config set model.base_url http://127.0.0.1:8080/v1 hermes config set model.default tahaalam2009/VeriLoop-E2-GSQ-RCO-GGUF:IQ2_S
Run Hermes
hermes
- Atomic Chat
- OpenClaw
How to use tahaalam2009/VeriLoop-E2-GSQ-RCO-GGUF with OpenClaw:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf tahaalam2009/VeriLoop-E2-GSQ-RCO-GGUF:IQ2_S
Configure OpenClaw
# Install OpenClaw: npm install -g openclaw@latest # Register the local server and set it as the default model: openclaw onboard --non-interactive --mode local \ --auth-choice custom-api-key \ --custom-base-url http://127.0.0.1:8080/v1 \ --custom-model-id "tahaalam2009/VeriLoop-E2-GSQ-RCO-GGUF:IQ2_S" \ --custom-provider-id llama-cpp \ --custom-compatibility openai \ --custom-text-input \ --accept-risk \ --skip-health
Run OpenClaw
openclaw agent --local --agent main --message "Hello from Hugging Face"
Upload eval_summary_deep_dive.json with huggingface_hub
Browse files- eval_summary_deep_dive.json +176 -265
eval_summary_deep_dive.json
CHANGED
|
@@ -2,331 +2,242 @@
|
|
| 2 |
"wikitext2": {
|
| 3 |
"name": "WikiText-2",
|
| 4 |
"bf16": {
|
| 5 |
-
"ppl": 5.
|
| 6 |
-
"ppl_unc": 0.
|
| 7 |
-
"time_sec": 195.8225395679474
|
| 8 |
},
|
| 9 |
"models": {
|
| 10 |
"VeriLoop_IQ3_XXS": {
|
| 11 |
-
"
|
| 12 |
-
"
|
| 13 |
-
"
|
| 14 |
-
"
|
| 15 |
-
"
|
| 16 |
-
"
|
| 17 |
-
"
|
| 18 |
-
"
|
| 19 |
-
"
|
| 20 |
-
"
|
| 21 |
-
"rms_delta_p": null,
|
| 22 |
-
"time_sec": 110.35492277145386
|
| 23 |
},
|
| 24 |
"VeriLoop_IQ3_S": {
|
| 25 |
-
"
|
| 26 |
-
"
|
| 27 |
-
"
|
| 28 |
-
"
|
| 29 |
-
"
|
| 30 |
-
"
|
| 31 |
-
"
|
| 32 |
-
"
|
| 33 |
-
"
|
| 34 |
-
"
|
| 35 |
-
"rms_delta_p": null,
|
| 36 |
-
"time_sec": 118.25298762321472
|
| 37 |
},
|
| 38 |
"Qwen_IQ3_XXS": {
|
| 39 |
-
"
|
| 40 |
-
"
|
| 41 |
-
"
|
| 42 |
-
"
|
| 43 |
-
"
|
| 44 |
-
"
|
| 45 |
-
"
|
| 46 |
-
"
|
| 47 |
-
"
|
| 48 |
-
"
|
| 49 |
-
"rms_delta_p": null,
|
| 50 |
-
"time_sec": 109.87469863891602
|
| 51 |
},
|
| 52 |
"Qwen_IQ3_S": {
|
| 53 |
-
"
|
| 54 |
-
"
|
| 55 |
-
"
|
| 56 |
-
"
|
| 57 |
-
"
|
| 58 |
-
"
|
| 59 |
-
"
|
| 60 |
-
"
|
| 61 |
-
"
|
| 62 |
-
"
|
| 63 |
-
"rms_delta_p": null,
|
| 64 |
-
"time_sec": 112.30604457855225
|
| 65 |
}
|
| 66 |
}
|
| 67 |
},
|
| 68 |
"codemath": {
|
| 69 |
"name": "Code & Math",
|
| 70 |
"bf16": {
|
| 71 |
-
"ppl": 1.
|
| 72 |
-
"ppl_unc": 0.
|
| 73 |
-
"time_sec": 148.9525682926178
|
| 74 |
},
|
| 75 |
"models": {
|
| 76 |
"VeriLoop_IQ3_XXS": {
|
| 77 |
-
"
|
| 78 |
-
"
|
| 79 |
-
"
|
| 80 |
-
"
|
| 81 |
-
"
|
| 82 |
-
"
|
| 83 |
-
"
|
| 84 |
-
"
|
| 85 |
-
"
|
| 86 |
-
"
|
| 87 |
-
"rms_delta_p": null,
|
| 88 |
-
"time_sec": 97.66222810745239
|
| 89 |
},
|
| 90 |
"VeriLoop_IQ3_S": {
|
| 91 |
-
"
|
| 92 |
-
"
|
| 93 |
-
"
|
| 94 |
-
"
|
| 95 |
-
"
|
| 96 |
-
"
|
| 97 |
-
"
|
| 98 |
-
"
|
| 99 |
-
"
|
| 100 |
-
"
|
| 101 |
-
"rms_delta_p": null,
|
| 102 |
-
"time_sec": 99.04169034957886
|
| 103 |
},
|
| 104 |
"Qwen_IQ3_XXS": {
|
| 105 |
-
"
|
| 106 |
-
"
|
| 107 |
-
"
|
| 108 |
-
"
|
| 109 |
-
"
|
| 110 |
-
"
|
| 111 |
-
"
|
| 112 |
-
"
|
| 113 |
-
"
|
| 114 |
-
"
|
| 115 |
-
"rms_delta_p": null,
|
| 116 |
-
"time_sec": 100.96973204612732
|
| 117 |
},
|
| 118 |
"Qwen_IQ3_S": {
|
| 119 |
-
"
|
| 120 |
-
"
|
| 121 |
-
"
|
| 122 |
-
"
|
| 123 |
-
"
|
| 124 |
-
"
|
| 125 |
-
"
|
| 126 |
-
"
|
| 127 |
-
"
|
| 128 |
-
"
|
| 129 |
-
"rms_delta_p": null,
|
| 130 |
-
"time_sec": 102.65990972518921
|
| 131 |
}
|
| 132 |
}
|
| 133 |
},
|
| 134 |
"aime2025": {
|
| 135 |
"name": "AIME 2025",
|
| 136 |
"bf16": {
|
| 137 |
-
"ppl": 2.
|
| 138 |
-
"ppl_unc": 0.
|
| 139 |
-
"time_sec": 153.58131432533264
|
| 140 |
},
|
| 141 |
"models": {
|
| 142 |
"VeriLoop_IQ3_XXS": {
|
| 143 |
-
"
|
| 144 |
-
"
|
| 145 |
-
"
|
| 146 |
-
"
|
| 147 |
-
"
|
| 148 |
-
"
|
| 149 |
-
"
|
| 150 |
-
"
|
| 151 |
-
"
|
| 152 |
-
"
|
| 153 |
-
"rms_delta_p": null,
|
| 154 |
-
"time_sec": 100.4238953590393
|
| 155 |
},
|
| 156 |
"VeriLoop_IQ3_S": {
|
| 157 |
-
"
|
| 158 |
-
"
|
| 159 |
-
"
|
| 160 |
-
"
|
| 161 |
-
"
|
| 162 |
-
"
|
| 163 |
-
"
|
| 164 |
-
"
|
| 165 |
-
"
|
| 166 |
-
"
|
| 167 |
-
"rms_delta_p": null,
|
| 168 |
-
"time_sec": 101.96496200561523
|
| 169 |
},
|
| 170 |
"Qwen_IQ3_XXS": {
|
| 171 |
-
"
|
| 172 |
-
"
|
| 173 |
-
"
|
| 174 |
-
"
|
| 175 |
-
"
|
| 176 |
-
"
|
| 177 |
-
"
|
| 178 |
-
"
|
| 179 |
-
"
|
| 180 |
-
"
|
| 181 |
-
"rms_delta_p": null,
|
| 182 |
-
"time_sec": 98.5217227935791
|
| 183 |
},
|
| 184 |
"Qwen_IQ3_S": {
|
| 185 |
-
"
|
| 186 |
-
"
|
| 187 |
-
"
|
| 188 |
-
"
|
| 189 |
-
"
|
| 190 |
-
"
|
| 191 |
-
"
|
| 192 |
-
"
|
| 193 |
-
"
|
| 194 |
-
"
|
| 195 |
-
"rms_delta_p": null,
|
| 196 |
-
"time_sec": 108.99933910369873
|
| 197 |
}
|
| 198 |
}
|
| 199 |
},
|
| 200 |
"livecode": {
|
| 201 |
"name": "LiveCodeBench",
|
| 202 |
"bf16": {
|
| 203 |
-
"ppl": 1.
|
| 204 |
-
"ppl_unc": 0.
|
| 205 |
-
"time_sec": 159.20217180252075
|
| 206 |
},
|
| 207 |
"models": {
|
| 208 |
"VeriLoop_IQ3_XXS": {
|
| 209 |
-
"
|
| 210 |
-
"
|
| 211 |
-
"
|
| 212 |
-
"
|
| 213 |
-
"
|
| 214 |
-
"
|
| 215 |
-
"
|
| 216 |
-
"
|
| 217 |
-
"
|
| 218 |
-
"
|
| 219 |
-
"rms_delta_p": null,
|
| 220 |
-
"time_sec": 101.15227603912354
|
| 221 |
},
|
| 222 |
"VeriLoop_IQ3_S": {
|
| 223 |
-
"
|
| 224 |
-
"
|
| 225 |
-
"
|
| 226 |
-
"
|
| 227 |
-
"
|
| 228 |
-
"
|
| 229 |
-
"
|
| 230 |
-
"
|
| 231 |
-
"
|
| 232 |
-
"
|
| 233 |
-
"rms_delta_p": null,
|
| 234 |
-
"time_sec": 98.8099889755249
|
| 235 |
},
|
| 236 |
"Qwen_IQ3_XXS": {
|
| 237 |
-
"
|
| 238 |
-
"
|
| 239 |
-
"
|
| 240 |
-
"
|
| 241 |
-
"
|
| 242 |
-
"
|
| 243 |
-
"
|
| 244 |
-
"
|
| 245 |
-
"
|
| 246 |
-
"
|
| 247 |
-
"rms_delta_p": null,
|
| 248 |
-
"time_sec": 98.93727707862854
|
| 249 |
},
|
| 250 |
"Qwen_IQ3_S": {
|
| 251 |
-
"
|
| 252 |
-
"
|
| 253 |
-
"
|
| 254 |
-
"
|
| 255 |
-
"
|
| 256 |
-
"
|
| 257 |
-
"
|
| 258 |
-
"
|
| 259 |
-
"
|
| 260 |
-
"
|
| 261 |
-
"rms_delta_p": null,
|
| 262 |
-
"time_sec": 97.8656849861145
|
| 263 |
}
|
| 264 |
}
|
| 265 |
},
|
| 266 |
"terminal": {
|
| 267 |
"name": "TerminalBench 2.1",
|
| 268 |
-
"bf16": {
|
| 269 |
-
"ppl": null,
|
| 270 |
-
"ppl_unc": null,
|
| 271 |
-
"time_sec": 17.898959398269653
|
| 272 |
-
},
|
| 273 |
"models": {
|
| 274 |
-
"VeriLoop_IQ3_XXS": {
|
| 275 |
-
|
| 276 |
-
|
| 277 |
-
|
| 278 |
-
|
| 279 |
-
|
| 280 |
-
|
| 281 |
-
|
| 282 |
-
"top1_unc": null,
|
| 283 |
-
"top5_agreement": null,
|
| 284 |
-
"top5_unc": null,
|
| 285 |
-
"rms_delta_p": null,
|
| 286 |
-
"time_sec": 3.4318151473999023
|
| 287 |
-
},
|
| 288 |
-
"VeriLoop_IQ3_S": {
|
| 289 |
-
"label": "VeriLoop-E2 GSQ-RCO IQ3_S (3.55 bpw)",
|
| 290 |
-
"ppl": null,
|
| 291 |
-
"ppl_unc": null,
|
| 292 |
-
"ppl_ratio": null,
|
| 293 |
-
"mean_kld": null,
|
| 294 |
-
"kld_unc": null,
|
| 295 |
-
"top1_agreement": null,
|
| 296 |
-
"top1_unc": null,
|
| 297 |
-
"top5_agreement": null,
|
| 298 |
-
"top5_unc": null,
|
| 299 |
-
"rms_delta_p": null,
|
| 300 |
-
"time_sec": 5.760640859603882
|
| 301 |
-
},
|
| 302 |
-
"Qwen_IQ3_XXS": {
|
| 303 |
-
"label": "Qwen3.8-27B GSQ-RCO IQ3_XXS (3.05 bpw)",
|
| 304 |
-
"ppl": null,
|
| 305 |
-
"ppl_unc": null,
|
| 306 |
-
"ppl_ratio": null,
|
| 307 |
-
"mean_kld": null,
|
| 308 |
-
"kld_unc": null,
|
| 309 |
-
"top1_agreement": null,
|
| 310 |
-
"top1_unc": null,
|
| 311 |
-
"top5_agreement": null,
|
| 312 |
-
"top5_unc": null,
|
| 313 |
-
"rms_delta_p": null,
|
| 314 |
-
"time_sec": 3.5282740592956543
|
| 315 |
-
},
|
| 316 |
-
"Qwen_IQ3_S": {
|
| 317 |
-
"label": "Qwen3.8-27B GSQ-RCO IQ3_S (3.55 bpw)",
|
| 318 |
-
"ppl": null,
|
| 319 |
-
"ppl_unc": null,
|
| 320 |
-
"ppl_ratio": null,
|
| 321 |
-
"mean_kld": null,
|
| 322 |
-
"kld_unc": null,
|
| 323 |
-
"top1_agreement": null,
|
| 324 |
-
"top1_unc": null,
|
| 325 |
-
"top5_agreement": null,
|
| 326 |
-
"top5_unc": null,
|
| 327 |
-
"rms_delta_p": null,
|
| 328 |
-
"time_sec": 4.589294672012329
|
| 329 |
-
}
|
| 330 |
}
|
| 331 |
}
|
| 332 |
}
|
|
|
|
| 2 |
"wikitext2": {
|
| 3 |
"name": "WikiText-2",
|
| 4 |
"bf16": {
|
| 5 |
+
"ppl": 5.157865,
|
| 6 |
+
"ppl_unc": 0.369376
|
|
|
|
| 7 |
},
|
| 8 |
"models": {
|
| 9 |
"VeriLoop_IQ3_XXS": {
|
| 10 |
+
"ppl": 5.281293,
|
| 11 |
+
"ppl_unc": 0.379551,
|
| 12 |
+
"ppl_ratio": 1.02393,
|
| 13 |
+
"mean_kld": 0.078914,
|
| 14 |
+
"kld_unc": 0.005409,
|
| 15 |
+
"top1": 88.748,
|
| 16 |
+
"top1_unc": 0.989,
|
| 17 |
+
"top5": 99.413,
|
| 18 |
+
"top5_unc": 0.239,
|
| 19 |
+
"rms": 7.67
|
|
|
|
|
|
|
| 20 |
},
|
| 21 |
"VeriLoop_IQ3_S": {
|
| 22 |
+
"ppl": 5.368311,
|
| 23 |
+
"ppl_unc": 0.392426,
|
| 24 |
+
"ppl_ratio": 1.040801,
|
| 25 |
+
"mean_kld": 0.035113,
|
| 26 |
+
"kld_unc": 0.001932,
|
| 27 |
+
"top1": 92.368,
|
| 28 |
+
"top1_unc": 0.831,
|
| 29 |
+
"top5": 99.804,
|
| 30 |
+
"top5_unc": 0.138,
|
| 31 |
+
"rms": 5.132
|
|
|
|
|
|
|
| 32 |
},
|
| 33 |
"Qwen_IQ3_XXS": {
|
| 34 |
+
"ppl": 5.391707,
|
| 35 |
+
"ppl_unc": 0.391388,
|
| 36 |
+
"ppl_ratio": 1.045337,
|
| 37 |
+
"mean_kld": 0.072731,
|
| 38 |
+
"kld_unc": 0.004163,
|
| 39 |
+
"top1": 89.335,
|
| 40 |
+
"top1_unc": 0.966,
|
| 41 |
+
"top5": 99.315,
|
| 42 |
+
"top5_unc": 0.258,
|
| 43 |
+
"rms": 7.338
|
|
|
|
|
|
|
| 44 |
},
|
| 45 |
"Qwen_IQ3_S": {
|
| 46 |
+
"ppl": 5.287517,
|
| 47 |
+
"ppl_unc": 0.383349,
|
| 48 |
+
"ppl_ratio": 1.025137,
|
| 49 |
+
"mean_kld": 0.043676,
|
| 50 |
+
"kld_unc": 0.003106,
|
| 51 |
+
"top1": 91.194,
|
| 52 |
+
"top1_unc": 0.887,
|
| 53 |
+
"top5": 99.706,
|
| 54 |
+
"top5_unc": 0.169,
|
| 55 |
+
"rms": 5.745
|
|
|
|
|
|
|
| 56 |
}
|
| 57 |
}
|
| 58 |
},
|
| 59 |
"codemath": {
|
| 60 |
"name": "Code & Math",
|
| 61 |
"bf16": {
|
| 62 |
+
"ppl": 1.670003,
|
| 63 |
+
"ppl_unc": 0.063744
|
|
|
|
| 64 |
},
|
| 65 |
"models": {
|
| 66 |
"VeriLoop_IQ3_XXS": {
|
| 67 |
+
"ppl": 2.129404,
|
| 68 |
+
"ppl_unc": 0.105036,
|
| 69 |
+
"ppl_ratio": 1.27509,
|
| 70 |
+
"mean_kld": 0.256974,
|
| 71 |
+
"kld_unc": 0.020245,
|
| 72 |
+
"top1": 86.106,
|
| 73 |
+
"top1_unc": 1.082,
|
| 74 |
+
"top5": 98.728,
|
| 75 |
+
"top5_unc": 0.351,
|
| 76 |
+
"rms": 22.225
|
|
|
|
|
|
|
| 77 |
},
|
| 78 |
"VeriLoop_IQ3_S": {
|
| 79 |
+
"ppl": 1.779099,
|
| 80 |
+
"ppl_unc": 0.068721,
|
| 81 |
+
"ppl_ratio": 1.065326,
|
| 82 |
+
"mean_kld": 0.126346,
|
| 83 |
+
"kld_unc": 0.013601,
|
| 84 |
+
"top1": 91.292,
|
| 85 |
+
"top1_unc": 0.882,
|
| 86 |
+
"top5": 99.511,
|
| 87 |
+
"top5_unc": 0.218,
|
| 88 |
+
"rms": 15.189
|
|
|
|
|
|
|
| 89 |
},
|
| 90 |
"Qwen_IQ3_XXS": {
|
| 91 |
+
"ppl": 2.604939,
|
| 92 |
+
"ppl_unc": 0.161942,
|
| 93 |
+
"ppl_ratio": 1.559841,
|
| 94 |
+
"mean_kld": 0.42651,
|
| 95 |
+
"kld_unc": 0.037084,
|
| 96 |
+
"top1": 82.485,
|
| 97 |
+
"top1_unc": 1.19,
|
| 98 |
+
"top5": 96.771,
|
| 99 |
+
"top5_unc": 0.553,
|
| 100 |
+
"rms": 26.362
|
|
|
|
|
|
|
| 101 |
},
|
| 102 |
"Qwen_IQ3_S": {
|
| 103 |
+
"ppl": 2.286893,
|
| 104 |
+
"ppl_unc": 0.128775,
|
| 105 |
+
"ppl_ratio": 1.369394,
|
| 106 |
+
"mean_kld": 0.27297,
|
| 107 |
+
"kld_unc": 0.023905,
|
| 108 |
+
"top1": 85.812,
|
| 109 |
+
"top1_unc": 1.092,
|
| 110 |
+
"top5": 98.337,
|
| 111 |
+
"top5_unc": 0.4,
|
| 112 |
+
"rms": 22.573
|
|
|
|
|
|
|
| 113 |
}
|
| 114 |
}
|
| 115 |
},
|
| 116 |
"aime2025": {
|
| 117 |
"name": "AIME 2025",
|
| 118 |
"bf16": {
|
| 119 |
+
"ppl": 2.073159,
|
| 120 |
+
"ppl_unc": 0.093297
|
|
|
|
| 121 |
},
|
| 122 |
"models": {
|
| 123 |
"VeriLoop_IQ3_XXS": {
|
| 124 |
+
"ppl": 2.142681,
|
| 125 |
+
"ppl_unc": 0.099012,
|
| 126 |
+
"ppl_ratio": 1.033534,
|
| 127 |
+
"mean_kld": 0.041285,
|
| 128 |
+
"kld_unc": 0.003125,
|
| 129 |
+
"top1": 94.227,
|
| 130 |
+
"top1_unc": 0.73,
|
| 131 |
+
"top5": 99.902,
|
| 132 |
+
"top5_unc": 0.098,
|
| 133 |
+
"rms": 7.302
|
|
|
|
|
|
|
| 134 |
},
|
| 135 |
"VeriLoop_IQ3_S": {
|
| 136 |
+
"ppl": 2.119914,
|
| 137 |
+
"ppl_unc": 0.097774,
|
| 138 |
+
"ppl_ratio": 1.022552,
|
| 139 |
+
"mean_kld": 0.025689,
|
| 140 |
+
"kld_unc": 0.001889,
|
| 141 |
+
"top1": 95.401,
|
| 142 |
+
"top1_unc": 0.656,
|
| 143 |
+
"top5": 99.804,
|
| 144 |
+
"top5_unc": 0.138,
|
| 145 |
+
"rms": 5.324
|
|
|
|
|
|
|
| 146 |
},
|
| 147 |
"Qwen_IQ3_XXS": {
|
| 148 |
+
"ppl": 2.196686,
|
| 149 |
+
"ppl_unc": 0.105841,
|
| 150 |
+
"ppl_ratio": 1.059584,
|
| 151 |
+
"mean_kld": 0.07322,
|
| 152 |
+
"kld_unc": 0.011335,
|
| 153 |
+
"top1": 93.151,
|
| 154 |
+
"top1_unc": 0.791,
|
| 155 |
+
"top5": 99.706,
|
| 156 |
+
"top5_unc": 0.169,
|
| 157 |
+
"rms": 10.09
|
|
|
|
|
|
|
| 158 |
},
|
| 159 |
"Qwen_IQ3_S": {
|
| 160 |
+
"ppl": 2.251226,
|
| 161 |
+
"ppl_unc": 0.112986,
|
| 162 |
+
"ppl_ratio": 1.085891,
|
| 163 |
+
"mean_kld": 0.090163,
|
| 164 |
+
"kld_unc": 0.014369,
|
| 165 |
+
"top1": 92.466,
|
| 166 |
+
"top1_unc": 0.826,
|
| 167 |
+
"top5": 99.413,
|
| 168 |
+
"top5_unc": 0.239,
|
| 169 |
+
"rms": 10.649
|
|
|
|
|
|
|
| 170 |
}
|
| 171 |
}
|
| 172 |
},
|
| 173 |
"livecode": {
|
| 174 |
"name": "LiveCodeBench",
|
| 175 |
"bf16": {
|
| 176 |
+
"ppl": 1.310318,
|
| 177 |
+
"ppl_unc": 0.044372
|
|
|
|
| 178 |
},
|
| 179 |
"models": {
|
| 180 |
"VeriLoop_IQ3_XXS": {
|
| 181 |
+
"ppl": 1.50444,
|
| 182 |
+
"ppl_unc": 0.058624,
|
| 183 |
+
"ppl_ratio": 1.148149,
|
| 184 |
+
"mean_kld": 0.249568,
|
| 185 |
+
"kld_unc": 0.027405,
|
| 186 |
+
"top1": 91.781,
|
| 187 |
+
"top1_unc": 0.86,
|
| 188 |
+
"top5": 98.826,
|
| 189 |
+
"top5_unc": 0.337,
|
| 190 |
+
"rms": 20.829
|
|
|
|
|
|
|
| 191 |
},
|
| 192 |
"VeriLoop_IQ3_S": {
|
| 193 |
+
"ppl": 1.393964,
|
| 194 |
+
"ppl_unc": 0.050473,
|
| 195 |
+
"ppl_ratio": 1.063836,
|
| 196 |
+
"mean_kld": 0.158601,
|
| 197 |
+
"kld_unc": 0.020799,
|
| 198 |
+
"top1": 94.031,
|
| 199 |
+
"top1_unc": 0.741,
|
| 200 |
+
"top5": 100.0,
|
| 201 |
+
"top5_unc": 0.0,
|
| 202 |
+
"rms": 16.79
|
|
|
|
|
|
|
| 203 |
},
|
| 204 |
"Qwen_IQ3_XXS": {
|
| 205 |
+
"ppl": 1.846662,
|
| 206 |
+
"ppl_unc": 0.10329,
|
| 207 |
+
"ppl_ratio": 1.409323,
|
| 208 |
+
"mean_kld": 0.38225,
|
| 209 |
+
"kld_unc": 0.039303,
|
| 210 |
+
"top1": 89.041,
|
| 211 |
+
"top1_unc": 0.978,
|
| 212 |
+
"top5": 98.239,
|
| 213 |
+
"top5_unc": 0.412,
|
| 214 |
+
"rms": 25.883
|
|
|
|
|
|
|
| 215 |
},
|
| 216 |
"Qwen_IQ3_S": {
|
| 217 |
+
"ppl": 1.633881,
|
| 218 |
+
"ppl_unc": 0.07884,
|
| 219 |
+
"ppl_ratio": 1.246935,
|
| 220 |
+
"mean_kld": 0.266109,
|
| 221 |
+
"kld_unc": 0.029915,
|
| 222 |
+
"top1": 91.096,
|
| 223 |
+
"top1_unc": 0.891,
|
| 224 |
+
"top5": 98.728,
|
| 225 |
+
"top5_unc": 0.351,
|
| 226 |
+
"rms": 22.114
|
|
|
|
|
|
|
| 227 |
}
|
| 228 |
}
|
| 229 |
},
|
| 230 |
"terminal": {
|
| 231 |
"name": "TerminalBench 2.1",
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 232 |
"models": {
|
| 233 |
+
"VeriLoop_IQ3_XXS": {},
|
| 234 |
+
"VeriLoop_IQ3_S": {},
|
| 235 |
+
"Qwen_IQ3_XXS": {},
|
| 236 |
+
"Qwen_IQ3_S": {}
|
| 237 |
+
},
|
| 238 |
+
"bf16": {
|
| 239 |
+
"ppl": 2.15,
|
| 240 |
+
"ppl_unc": 0.08
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 241 |
}
|
| 242 |
}
|
| 243 |
}
|