Image-Text-to-Text
Transformers
Safetensors
qwen3_5
qwen3.8
fp8
compressed-tensors
multimodal
vision
reasoning
swissneuron
conversational
Instructions to use SwissNeuron/Qwen3.8-27B-SwissNeuron-Derisked-FP8 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use SwissNeuron/Qwen3.8-27B-SwissNeuron-Derisked-FP8 with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("image-text-to-text", model="SwissNeuron/Qwen3.8-27B-SwissNeuron-Derisked-FP8") messages = [ { "role": "user", "content": [ {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"}, {"type": "text", "text": "What animal is on the candy?"} ] }, ] pipe(text=messages)# Load model directly from transformers import AutoProcessor, AutoModelForMultimodalLM processor = AutoProcessor.from_pretrained("SwissNeuron/Qwen3.8-27B-SwissNeuron-Derisked-FP8") model = AutoModelForMultimodalLM.from_pretrained("SwissNeuron/Qwen3.8-27B-SwissNeuron-Derisked-FP8", device_map="auto") messages = [ { "role": "user", "content": [ {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"}, {"type": "text", "text": "What animal is on the candy?"} ] }, ] inputs = processor.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(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:])) - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use SwissNeuron/Qwen3.8-27B-SwissNeuron-Derisked-FP8 with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "SwissNeuron/Qwen3.8-27B-SwissNeuron-Derisked-FP8" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "SwissNeuron/Qwen3.8-27B-SwissNeuron-Derisked-FP8", "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/SwissNeuron/Qwen3.8-27B-SwissNeuron-Derisked-FP8
- SGLang
How to use SwissNeuron/Qwen3.8-27B-SwissNeuron-Derisked-FP8 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 "SwissNeuron/Qwen3.8-27B-SwissNeuron-Derisked-FP8" \ --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": "SwissNeuron/Qwen3.8-27B-SwissNeuron-Derisked-FP8", "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 "SwissNeuron/Qwen3.8-27B-SwissNeuron-Derisked-FP8" \ --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": "SwissNeuron/Qwen3.8-27B-SwissNeuron-Derisked-FP8", "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 SwissNeuron/Qwen3.8-27B-SwissNeuron-Derisked-FP8 with Docker Model Runner:
docker model run hf.co/SwissNeuron/Qwen3.8-27B-SwissNeuron-Derisked-FP8
Exclude restored MTP from FP8 quantization
Browse files- config.json +139 -124
config.json
CHANGED
|
@@ -49,132 +49,11 @@
|
|
| 49 |
"format": "float-quantized",
|
| 50 |
"global_compression_ratio": null,
|
| 51 |
"ignore": [
|
| 52 |
-
"
|
| 53 |
-
"model.visual.blocks.0.attn.proj",
|
| 54 |
-
"model.visual.blocks.0.mlp.linear_fc1",
|
| 55 |
-
"model.visual.blocks.0.mlp.linear_fc2",
|
| 56 |
-
"model.visual.blocks.1.attn.qkv",
|
| 57 |
-
"model.visual.blocks.1.attn.proj",
|
| 58 |
-
"model.visual.blocks.1.mlp.linear_fc1",
|
| 59 |
-
"model.visual.blocks.1.mlp.linear_fc2",
|
| 60 |
-
"model.visual.blocks.2.attn.qkv",
|
| 61 |
-
"model.visual.blocks.2.attn.proj",
|
| 62 |
-
"model.visual.blocks.2.mlp.linear_fc1",
|
| 63 |
-
"model.visual.blocks.2.mlp.linear_fc2",
|
| 64 |
-
"model.visual.blocks.3.attn.qkv",
|
| 65 |
-
"model.visual.blocks.3.attn.proj",
|
| 66 |
-
"model.visual.blocks.3.mlp.linear_fc1",
|
| 67 |
-
"model.visual.blocks.3.mlp.linear_fc2",
|
| 68 |
-
"model.visual.blocks.4.attn.qkv",
|
| 69 |
-
"model.visual.blocks.4.attn.proj",
|
| 70 |
-
"model.visual.blocks.4.mlp.linear_fc1",
|
| 71 |
-
"model.visual.blocks.4.mlp.linear_fc2",
|
| 72 |
-
"model.visual.blocks.5.attn.qkv",
|
| 73 |
-
"model.visual.blocks.5.attn.proj",
|
| 74 |
-
"model.visual.blocks.5.mlp.linear_fc1",
|
| 75 |
-
"model.visual.blocks.5.mlp.linear_fc2",
|
| 76 |
-
"model.visual.blocks.6.attn.qkv",
|
| 77 |
-
"model.visual.blocks.6.attn.proj",
|
| 78 |
-
"model.visual.blocks.6.mlp.linear_fc1",
|
| 79 |
-
"model.visual.blocks.6.mlp.linear_fc2",
|
| 80 |
-
"model.visual.blocks.7.attn.qkv",
|
| 81 |
-
"model.visual.blocks.7.attn.proj",
|
| 82 |
-
"model.visual.blocks.7.mlp.linear_fc1",
|
| 83 |
-
"model.visual.blocks.7.mlp.linear_fc2",
|
| 84 |
-
"model.visual.blocks.8.attn.qkv",
|
| 85 |
-
"model.visual.blocks.8.attn.proj",
|
| 86 |
-
"model.visual.blocks.8.mlp.linear_fc1",
|
| 87 |
-
"model.visual.blocks.8.mlp.linear_fc2",
|
| 88 |
-
"model.visual.blocks.9.attn.qkv",
|
| 89 |
-
"model.visual.blocks.9.attn.proj",
|
| 90 |
-
"model.visual.blocks.9.mlp.linear_fc1",
|
| 91 |
-
"model.visual.blocks.9.mlp.linear_fc2",
|
| 92 |
-
"model.visual.blocks.10.attn.qkv",
|
| 93 |
-
"model.visual.blocks.10.attn.proj",
|
| 94 |
-
"model.visual.blocks.10.mlp.linear_fc1",
|
| 95 |
-
"model.visual.blocks.10.mlp.linear_fc2",
|
| 96 |
-
"model.visual.blocks.11.attn.qkv",
|
| 97 |
-
"model.visual.blocks.11.attn.proj",
|
| 98 |
-
"model.visual.blocks.11.mlp.linear_fc1",
|
| 99 |
-
"model.visual.blocks.11.mlp.linear_fc2",
|
| 100 |
-
"model.visual.blocks.12.attn.qkv",
|
| 101 |
-
"model.visual.blocks.12.attn.proj",
|
| 102 |
-
"model.visual.blocks.12.mlp.linear_fc1",
|
| 103 |
-
"model.visual.blocks.12.mlp.linear_fc2",
|
| 104 |
-
"model.visual.blocks.13.attn.qkv",
|
| 105 |
-
"model.visual.blocks.13.attn.proj",
|
| 106 |
-
"model.visual.blocks.13.mlp.linear_fc1",
|
| 107 |
-
"model.visual.blocks.13.mlp.linear_fc2",
|
| 108 |
-
"model.visual.blocks.14.attn.qkv",
|
| 109 |
-
"model.visual.blocks.14.attn.proj",
|
| 110 |
-
"model.visual.blocks.14.mlp.linear_fc1",
|
| 111 |
-
"model.visual.blocks.14.mlp.linear_fc2",
|
| 112 |
-
"model.visual.blocks.15.attn.qkv",
|
| 113 |
-
"model.visual.blocks.15.attn.proj",
|
| 114 |
-
"model.visual.blocks.15.mlp.linear_fc1",
|
| 115 |
-
"model.visual.blocks.15.mlp.linear_fc2",
|
| 116 |
-
"model.visual.blocks.16.attn.qkv",
|
| 117 |
-
"model.visual.blocks.16.attn.proj",
|
| 118 |
-
"model.visual.blocks.16.mlp.linear_fc1",
|
| 119 |
-
"model.visual.blocks.16.mlp.linear_fc2",
|
| 120 |
-
"model.visual.blocks.17.attn.qkv",
|
| 121 |
-
"model.visual.blocks.17.attn.proj",
|
| 122 |
-
"model.visual.blocks.17.mlp.linear_fc1",
|
| 123 |
-
"model.visual.blocks.17.mlp.linear_fc2",
|
| 124 |
-
"model.visual.blocks.18.attn.qkv",
|
| 125 |
-
"model.visual.blocks.18.attn.proj",
|
| 126 |
-
"model.visual.blocks.18.mlp.linear_fc1",
|
| 127 |
-
"model.visual.blocks.18.mlp.linear_fc2",
|
| 128 |
-
"model.visual.blocks.19.attn.qkv",
|
| 129 |
-
"model.visual.blocks.19.attn.proj",
|
| 130 |
-
"model.visual.blocks.19.mlp.linear_fc1",
|
| 131 |
-
"model.visual.blocks.19.mlp.linear_fc2",
|
| 132 |
-
"model.visual.blocks.20.attn.qkv",
|
| 133 |
-
"model.visual.blocks.20.attn.proj",
|
| 134 |
-
"model.visual.blocks.20.mlp.linear_fc1",
|
| 135 |
-
"model.visual.blocks.20.mlp.linear_fc2",
|
| 136 |
-
"model.visual.blocks.21.attn.qkv",
|
| 137 |
-
"model.visual.blocks.21.attn.proj",
|
| 138 |
-
"model.visual.blocks.21.mlp.linear_fc1",
|
| 139 |
-
"model.visual.blocks.21.mlp.linear_fc2",
|
| 140 |
-
"model.visual.blocks.22.attn.qkv",
|
| 141 |
-
"model.visual.blocks.22.attn.proj",
|
| 142 |
-
"model.visual.blocks.22.mlp.linear_fc1",
|
| 143 |
-
"model.visual.blocks.22.mlp.linear_fc2",
|
| 144 |
-
"model.visual.blocks.23.attn.qkv",
|
| 145 |
-
"model.visual.blocks.23.attn.proj",
|
| 146 |
-
"model.visual.blocks.23.mlp.linear_fc1",
|
| 147 |
-
"model.visual.blocks.23.mlp.linear_fc2",
|
| 148 |
-
"model.visual.blocks.24.attn.qkv",
|
| 149 |
-
"model.visual.blocks.24.attn.proj",
|
| 150 |
-
"model.visual.blocks.24.mlp.linear_fc1",
|
| 151 |
-
"model.visual.blocks.24.mlp.linear_fc2",
|
| 152 |
-
"model.visual.blocks.25.attn.qkv",
|
| 153 |
-
"model.visual.blocks.25.attn.proj",
|
| 154 |
-
"model.visual.blocks.25.mlp.linear_fc1",
|
| 155 |
-
"model.visual.blocks.25.mlp.linear_fc2",
|
| 156 |
-
"model.visual.blocks.26.attn.qkv",
|
| 157 |
-
"model.visual.blocks.26.attn.proj",
|
| 158 |
-
"model.visual.blocks.26.mlp.linear_fc1",
|
| 159 |
-
"model.visual.blocks.26.mlp.linear_fc2",
|
| 160 |
-
"model.visual.merger.linear_fc1",
|
| 161 |
-
"model.visual.merger.linear_fc2",
|
| 162 |
"model.language_model.layers.0.linear_attn",
|
| 163 |
"model.language_model.layers.0.linear_attn.norm",
|
| 164 |
"model.language_model.layers.1.linear_attn",
|
| 165 |
"model.language_model.layers.1.linear_attn.norm",
|
| 166 |
-
"model.language_model.layers.2.linear_attn",
|
| 167 |
-
"model.language_model.layers.2.linear_attn.norm",
|
| 168 |
-
"model.language_model.layers.4.linear_attn",
|
| 169 |
-
"model.language_model.layers.4.linear_attn.norm",
|
| 170 |
-
"model.language_model.layers.5.linear_attn",
|
| 171 |
-
"model.language_model.layers.5.linear_attn.norm",
|
| 172 |
-
"model.language_model.layers.6.linear_attn",
|
| 173 |
-
"model.language_model.layers.6.linear_attn.norm",
|
| 174 |
-
"model.language_model.layers.8.linear_attn",
|
| 175 |
-
"model.language_model.layers.8.linear_attn.norm",
|
| 176 |
-
"model.language_model.layers.9.linear_attn",
|
| 177 |
-
"model.language_model.layers.9.linear_attn.norm",
|
| 178 |
"model.language_model.layers.10.linear_attn",
|
| 179 |
"model.language_model.layers.10.linear_attn.norm",
|
| 180 |
"model.language_model.layers.12.linear_attn",
|
|
@@ -189,6 +68,8 @@
|
|
| 189 |
"model.language_model.layers.17.linear_attn.norm",
|
| 190 |
"model.language_model.layers.18.linear_attn",
|
| 191 |
"model.language_model.layers.18.linear_attn.norm",
|
|
|
|
|
|
|
| 192 |
"model.language_model.layers.20.linear_attn",
|
| 193 |
"model.language_model.layers.20.linear_attn.norm",
|
| 194 |
"model.language_model.layers.21.linear_attn",
|
|
@@ -219,6 +100,8 @@
|
|
| 219 |
"model.language_model.layers.37.linear_attn.norm",
|
| 220 |
"model.language_model.layers.38.linear_attn",
|
| 221 |
"model.language_model.layers.38.linear_attn.norm",
|
|
|
|
|
|
|
| 222 |
"model.language_model.layers.40.linear_attn",
|
| 223 |
"model.language_model.layers.40.linear_attn.norm",
|
| 224 |
"model.language_model.layers.41.linear_attn",
|
|
@@ -235,6 +118,8 @@
|
|
| 235 |
"model.language_model.layers.48.linear_attn.norm",
|
| 236 |
"model.language_model.layers.49.linear_attn",
|
| 237 |
"model.language_model.layers.49.linear_attn.norm",
|
|
|
|
|
|
|
| 238 |
"model.language_model.layers.50.linear_attn",
|
| 239 |
"model.language_model.layers.50.linear_attn.norm",
|
| 240 |
"model.language_model.layers.52.linear_attn",
|
|
@@ -249,13 +134,143 @@
|
|
| 249 |
"model.language_model.layers.57.linear_attn.norm",
|
| 250 |
"model.language_model.layers.58.linear_attn",
|
| 251 |
"model.language_model.layers.58.linear_attn.norm",
|
|
|
|
|
|
|
| 252 |
"model.language_model.layers.60.linear_attn",
|
| 253 |
"model.language_model.layers.60.linear_attn.norm",
|
| 254 |
"model.language_model.layers.61.linear_attn",
|
| 255 |
"model.language_model.layers.61.linear_attn.norm",
|
| 256 |
"model.language_model.layers.62.linear_attn",
|
| 257 |
"model.language_model.layers.62.linear_attn.norm",
|
| 258 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 259 |
],
|
| 260 |
"kv_cache_scheme": null,
|
| 261 |
"quant_method": "compressed-tensors",
|
|
@@ -406,4 +421,4 @@
|
|
| 406 |
},
|
| 407 |
"vision_end_token_id": 248054,
|
| 408 |
"vision_start_token_id": 248053
|
| 409 |
-
}
|
|
|
|
| 49 |
"format": "float-quantized",
|
| 50 |
"global_compression_ratio": null,
|
| 51 |
"ignore": [
|
| 52 |
+
"lm_head",
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 53 |
"model.language_model.layers.0.linear_attn",
|
| 54 |
"model.language_model.layers.0.linear_attn.norm",
|
| 55 |
"model.language_model.layers.1.linear_attn",
|
| 56 |
"model.language_model.layers.1.linear_attn.norm",
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 57 |
"model.language_model.layers.10.linear_attn",
|
| 58 |
"model.language_model.layers.10.linear_attn.norm",
|
| 59 |
"model.language_model.layers.12.linear_attn",
|
|
|
|
| 68 |
"model.language_model.layers.17.linear_attn.norm",
|
| 69 |
"model.language_model.layers.18.linear_attn",
|
| 70 |
"model.language_model.layers.18.linear_attn.norm",
|
| 71 |
+
"model.language_model.layers.2.linear_attn",
|
| 72 |
+
"model.language_model.layers.2.linear_attn.norm",
|
| 73 |
"model.language_model.layers.20.linear_attn",
|
| 74 |
"model.language_model.layers.20.linear_attn.norm",
|
| 75 |
"model.language_model.layers.21.linear_attn",
|
|
|
|
| 100 |
"model.language_model.layers.37.linear_attn.norm",
|
| 101 |
"model.language_model.layers.38.linear_attn",
|
| 102 |
"model.language_model.layers.38.linear_attn.norm",
|
| 103 |
+
"model.language_model.layers.4.linear_attn",
|
| 104 |
+
"model.language_model.layers.4.linear_attn.norm",
|
| 105 |
"model.language_model.layers.40.linear_attn",
|
| 106 |
"model.language_model.layers.40.linear_attn.norm",
|
| 107 |
"model.language_model.layers.41.linear_attn",
|
|
|
|
| 118 |
"model.language_model.layers.48.linear_attn.norm",
|
| 119 |
"model.language_model.layers.49.linear_attn",
|
| 120 |
"model.language_model.layers.49.linear_attn.norm",
|
| 121 |
+
"model.language_model.layers.5.linear_attn",
|
| 122 |
+
"model.language_model.layers.5.linear_attn.norm",
|
| 123 |
"model.language_model.layers.50.linear_attn",
|
| 124 |
"model.language_model.layers.50.linear_attn.norm",
|
| 125 |
"model.language_model.layers.52.linear_attn",
|
|
|
|
| 134 |
"model.language_model.layers.57.linear_attn.norm",
|
| 135 |
"model.language_model.layers.58.linear_attn",
|
| 136 |
"model.language_model.layers.58.linear_attn.norm",
|
| 137 |
+
"model.language_model.layers.6.linear_attn",
|
| 138 |
+
"model.language_model.layers.6.linear_attn.norm",
|
| 139 |
"model.language_model.layers.60.linear_attn",
|
| 140 |
"model.language_model.layers.60.linear_attn.norm",
|
| 141 |
"model.language_model.layers.61.linear_attn",
|
| 142 |
"model.language_model.layers.61.linear_attn.norm",
|
| 143 |
"model.language_model.layers.62.linear_attn",
|
| 144 |
"model.language_model.layers.62.linear_attn.norm",
|
| 145 |
+
"model.language_model.layers.8.linear_attn",
|
| 146 |
+
"model.language_model.layers.8.linear_attn.norm",
|
| 147 |
+
"model.language_model.layers.9.linear_attn",
|
| 148 |
+
"model.language_model.layers.9.linear_attn.norm",
|
| 149 |
+
"model.visual.blocks.0.attn.proj",
|
| 150 |
+
"model.visual.blocks.0.attn.qkv",
|
| 151 |
+
"model.visual.blocks.0.mlp.linear_fc1",
|
| 152 |
+
"model.visual.blocks.0.mlp.linear_fc2",
|
| 153 |
+
"model.visual.blocks.1.attn.proj",
|
| 154 |
+
"model.visual.blocks.1.attn.qkv",
|
| 155 |
+
"model.visual.blocks.1.mlp.linear_fc1",
|
| 156 |
+
"model.visual.blocks.1.mlp.linear_fc2",
|
| 157 |
+
"model.visual.blocks.10.attn.proj",
|
| 158 |
+
"model.visual.blocks.10.attn.qkv",
|
| 159 |
+
"model.visual.blocks.10.mlp.linear_fc1",
|
| 160 |
+
"model.visual.blocks.10.mlp.linear_fc2",
|
| 161 |
+
"model.visual.blocks.11.attn.proj",
|
| 162 |
+
"model.visual.blocks.11.attn.qkv",
|
| 163 |
+
"model.visual.blocks.11.mlp.linear_fc1",
|
| 164 |
+
"model.visual.blocks.11.mlp.linear_fc2",
|
| 165 |
+
"model.visual.blocks.12.attn.proj",
|
| 166 |
+
"model.visual.blocks.12.attn.qkv",
|
| 167 |
+
"model.visual.blocks.12.mlp.linear_fc1",
|
| 168 |
+
"model.visual.blocks.12.mlp.linear_fc2",
|
| 169 |
+
"model.visual.blocks.13.attn.proj",
|
| 170 |
+
"model.visual.blocks.13.attn.qkv",
|
| 171 |
+
"model.visual.blocks.13.mlp.linear_fc1",
|
| 172 |
+
"model.visual.blocks.13.mlp.linear_fc2",
|
| 173 |
+
"model.visual.blocks.14.attn.proj",
|
| 174 |
+
"model.visual.blocks.14.attn.qkv",
|
| 175 |
+
"model.visual.blocks.14.mlp.linear_fc1",
|
| 176 |
+
"model.visual.blocks.14.mlp.linear_fc2",
|
| 177 |
+
"model.visual.blocks.15.attn.proj",
|
| 178 |
+
"model.visual.blocks.15.attn.qkv",
|
| 179 |
+
"model.visual.blocks.15.mlp.linear_fc1",
|
| 180 |
+
"model.visual.blocks.15.mlp.linear_fc2",
|
| 181 |
+
"model.visual.blocks.16.attn.proj",
|
| 182 |
+
"model.visual.blocks.16.attn.qkv",
|
| 183 |
+
"model.visual.blocks.16.mlp.linear_fc1",
|
| 184 |
+
"model.visual.blocks.16.mlp.linear_fc2",
|
| 185 |
+
"model.visual.blocks.17.attn.proj",
|
| 186 |
+
"model.visual.blocks.17.attn.qkv",
|
| 187 |
+
"model.visual.blocks.17.mlp.linear_fc1",
|
| 188 |
+
"model.visual.blocks.17.mlp.linear_fc2",
|
| 189 |
+
"model.visual.blocks.18.attn.proj",
|
| 190 |
+
"model.visual.blocks.18.attn.qkv",
|
| 191 |
+
"model.visual.blocks.18.mlp.linear_fc1",
|
| 192 |
+
"model.visual.blocks.18.mlp.linear_fc2",
|
| 193 |
+
"model.visual.blocks.19.attn.proj",
|
| 194 |
+
"model.visual.blocks.19.attn.qkv",
|
| 195 |
+
"model.visual.blocks.19.mlp.linear_fc1",
|
| 196 |
+
"model.visual.blocks.19.mlp.linear_fc2",
|
| 197 |
+
"model.visual.blocks.2.attn.proj",
|
| 198 |
+
"model.visual.blocks.2.attn.qkv",
|
| 199 |
+
"model.visual.blocks.2.mlp.linear_fc1",
|
| 200 |
+
"model.visual.blocks.2.mlp.linear_fc2",
|
| 201 |
+
"model.visual.blocks.20.attn.proj",
|
| 202 |
+
"model.visual.blocks.20.attn.qkv",
|
| 203 |
+
"model.visual.blocks.20.mlp.linear_fc1",
|
| 204 |
+
"model.visual.blocks.20.mlp.linear_fc2",
|
| 205 |
+
"model.visual.blocks.21.attn.proj",
|
| 206 |
+
"model.visual.blocks.21.attn.qkv",
|
| 207 |
+
"model.visual.blocks.21.mlp.linear_fc1",
|
| 208 |
+
"model.visual.blocks.21.mlp.linear_fc2",
|
| 209 |
+
"model.visual.blocks.22.attn.proj",
|
| 210 |
+
"model.visual.blocks.22.attn.qkv",
|
| 211 |
+
"model.visual.blocks.22.mlp.linear_fc1",
|
| 212 |
+
"model.visual.blocks.22.mlp.linear_fc2",
|
| 213 |
+
"model.visual.blocks.23.attn.proj",
|
| 214 |
+
"model.visual.blocks.23.attn.qkv",
|
| 215 |
+
"model.visual.blocks.23.mlp.linear_fc1",
|
| 216 |
+
"model.visual.blocks.23.mlp.linear_fc2",
|
| 217 |
+
"model.visual.blocks.24.attn.proj",
|
| 218 |
+
"model.visual.blocks.24.attn.qkv",
|
| 219 |
+
"model.visual.blocks.24.mlp.linear_fc1",
|
| 220 |
+
"model.visual.blocks.24.mlp.linear_fc2",
|
| 221 |
+
"model.visual.blocks.25.attn.proj",
|
| 222 |
+
"model.visual.blocks.25.attn.qkv",
|
| 223 |
+
"model.visual.blocks.25.mlp.linear_fc1",
|
| 224 |
+
"model.visual.blocks.25.mlp.linear_fc2",
|
| 225 |
+
"model.visual.blocks.26.attn.proj",
|
| 226 |
+
"model.visual.blocks.26.attn.qkv",
|
| 227 |
+
"model.visual.blocks.26.mlp.linear_fc1",
|
| 228 |
+
"model.visual.blocks.26.mlp.linear_fc2",
|
| 229 |
+
"model.visual.blocks.3.attn.proj",
|
| 230 |
+
"model.visual.blocks.3.attn.qkv",
|
| 231 |
+
"model.visual.blocks.3.mlp.linear_fc1",
|
| 232 |
+
"model.visual.blocks.3.mlp.linear_fc2",
|
| 233 |
+
"model.visual.blocks.4.attn.proj",
|
| 234 |
+
"model.visual.blocks.4.attn.qkv",
|
| 235 |
+
"model.visual.blocks.4.mlp.linear_fc1",
|
| 236 |
+
"model.visual.blocks.4.mlp.linear_fc2",
|
| 237 |
+
"model.visual.blocks.5.attn.proj",
|
| 238 |
+
"model.visual.blocks.5.attn.qkv",
|
| 239 |
+
"model.visual.blocks.5.mlp.linear_fc1",
|
| 240 |
+
"model.visual.blocks.5.mlp.linear_fc2",
|
| 241 |
+
"model.visual.blocks.6.attn.proj",
|
| 242 |
+
"model.visual.blocks.6.attn.qkv",
|
| 243 |
+
"model.visual.blocks.6.mlp.linear_fc1",
|
| 244 |
+
"model.visual.blocks.6.mlp.linear_fc2",
|
| 245 |
+
"model.visual.blocks.7.attn.proj",
|
| 246 |
+
"model.visual.blocks.7.attn.qkv",
|
| 247 |
+
"model.visual.blocks.7.mlp.linear_fc1",
|
| 248 |
+
"model.visual.blocks.7.mlp.linear_fc2",
|
| 249 |
+
"model.visual.blocks.8.attn.proj",
|
| 250 |
+
"model.visual.blocks.8.attn.qkv",
|
| 251 |
+
"model.visual.blocks.8.mlp.linear_fc1",
|
| 252 |
+
"model.visual.blocks.8.mlp.linear_fc2",
|
| 253 |
+
"model.visual.blocks.9.attn.proj",
|
| 254 |
+
"model.visual.blocks.9.attn.qkv",
|
| 255 |
+
"model.visual.blocks.9.mlp.linear_fc1",
|
| 256 |
+
"model.visual.blocks.9.mlp.linear_fc2",
|
| 257 |
+
"model.visual.merger.linear_fc1",
|
| 258 |
+
"model.visual.merger.linear_fc2",
|
| 259 |
+
"mtp.fc",
|
| 260 |
+
"mtp.layers.0.input_layernorm",
|
| 261 |
+
"mtp.layers.0.mlp.down_proj",
|
| 262 |
+
"mtp.layers.0.mlp.gate_proj",
|
| 263 |
+
"mtp.layers.0.mlp.up_proj",
|
| 264 |
+
"mtp.layers.0.post_attention_layernorm",
|
| 265 |
+
"mtp.layers.0.self_attn.k_norm",
|
| 266 |
+
"mtp.layers.0.self_attn.k_proj",
|
| 267 |
+
"mtp.layers.0.self_attn.o_proj",
|
| 268 |
+
"mtp.layers.0.self_attn.q_norm",
|
| 269 |
+
"mtp.layers.0.self_attn.q_proj",
|
| 270 |
+
"mtp.layers.0.self_attn.v_proj",
|
| 271 |
+
"mtp.norm",
|
| 272 |
+
"mtp.pre_fc_norm_embedding",
|
| 273 |
+
"mtp.pre_fc_norm_hidden"
|
| 274 |
],
|
| 275 |
"kv_cache_scheme": null,
|
| 276 |
"quant_method": "compressed-tensors",
|
|
|
|
| 421 |
},
|
| 422 |
"vision_end_token_id": 248054,
|
| 423 |
"vision_start_token_id": 248053
|
| 424 |
+
}
|