Image-Text-to-Text
Transformers
Safetensors
minimax_m3_vl
multimodal
Mixture of Experts
agent
coding
video
conversational
custom_code
8-bit precision
quark
Instructions to use amd/MiniMax-M3-MXFP4 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use amd/MiniMax-M3-MXFP4 with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("image-text-to-text", model="amd/MiniMax-M3-MXFP4", trust_remote_code=True) 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("amd/MiniMax-M3-MXFP4", trust_remote_code=True) model = AutoModelForMultimodalLM.from_pretrained("amd/MiniMax-M3-MXFP4", trust_remote_code=True) 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 amd/MiniMax-M3-MXFP4 with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "amd/MiniMax-M3-MXFP4" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "amd/MiniMax-M3-MXFP4", "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/amd/MiniMax-M3-MXFP4
- SGLang
How to use amd/MiniMax-M3-MXFP4 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 "amd/MiniMax-M3-MXFP4" \ --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": "amd/MiniMax-M3-MXFP4", "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 "amd/MiniMax-M3-MXFP4" \ --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": "amd/MiniMax-M3-MXFP4", "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 amd/MiniMax-M3-MXFP4 with Docker Model Runner:
docker model run hf.co/amd/MiniMax-M3-MXFP4
| { | |
| "architectures": [ | |
| "MiniMaxM3SparseForConditionalGeneration" | |
| ], | |
| "auto_map": { | |
| "AutoConfig": "configuration_minimax_m3_vl.MiniMaxM3VLConfig" | |
| }, | |
| "model_type": "minimax_m3_vl", | |
| "text_config": { | |
| "hidden_size": 6144, | |
| "intermediate_size": 3072, | |
| "num_hidden_layers": 60, | |
| "num_attention_heads": 64, | |
| "num_key_value_heads": 4, | |
| "head_dim": 128, | |
| "vocab_size": 200064, | |
| "max_position_embeddings": 1048576, | |
| "rms_norm_eps": 1e-06, | |
| "use_gemma_norm": true, | |
| "attention_output_gate": false, | |
| "rope_theta": 5000000, | |
| "rotary_dim": 64, | |
| "partial_rotary_factor": 0.5, | |
| "hidden_act": "swigluoai", | |
| "use_qk_norm": true, | |
| "tie_word_embeddings": false, | |
| "dense_intermediate_size": 12288, | |
| "shared_intermediate_size": 3072, | |
| "num_local_experts": 128, | |
| "num_experts_per_tok": 4, | |
| "n_shared_experts": 1, | |
| "scoring_func": "sigmoid", | |
| "use_routing_bias": true, | |
| "moe_layer_freq": [ | |
| 0, | |
| 0, | |
| 0, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1 | |
| ], | |
| "qk_norm_type": "per_head", | |
| "num_mtp_modules": 7, | |
| "num_nextn_predict_layers": 1, | |
| "swiglu_alpha": 1.702, | |
| "swiglu_limit": 7.0, | |
| "routed_scaling_factor": 2.0, | |
| "sparse_attention_config": { | |
| "use_sparse_attention": true, | |
| "sparse_index_dim": 128, | |
| "sparse_num_index_heads": 4, | |
| "sparse_topk_blocks": 16, | |
| "sparse_block_size": 128, | |
| "sparse_disable_index_value": [ | |
| 0, | |
| 0, | |
| 0, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1 | |
| ], | |
| "sparse_score_type": "max", | |
| "sparse_init_block": 0, | |
| "sparse_local_block": 1, | |
| "sparse_attention_freq": [ | |
| 0, | |
| 0, | |
| 0, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1, | |
| 1 | |
| ] | |
| }, | |
| "architectures": [ | |
| "MiniMaxM3SparseForCausalLM" | |
| ] | |
| }, | |
| "vision_config": { | |
| "hidden_size": 1280, | |
| "num_attention_heads": 16, | |
| "num_hidden_layers": 32, | |
| "intermediate_size": 5120, | |
| "patch_size": 14, | |
| "image_size": 2016, | |
| "projection_dim": 6144, | |
| "position_embedding_type": "rope", | |
| "rope_mode": "3d", | |
| "rope_theta": 10000.0, | |
| "attention_dropout": 0.0, | |
| "hidden_act": "gelu", | |
| "initializer_factor": 1.0, | |
| "initializer_range": 0.02, | |
| "layer_norm_eps": 1e-05, | |
| "model_type": "clip_vision_model", | |
| "num_channels": 3, | |
| "vocab_size": 32000, | |
| "img_token_compression_config": { | |
| "image_token_compression_method": "patch_merge", | |
| "spatial_merge_size": 2, | |
| "temporal_patch_size": 2 | |
| }, | |
| "vision_segment_max_frames": 4 | |
| }, | |
| "img_token_compression_config": { | |
| "image_token_compression_method": "patch_merge", | |
| "spatial_merge_size": 2, | |
| "temporal_patch_size": 2 | |
| }, | |
| "image_grid_pinpoints": "[(336, 336), (336, 672), (336, 1008), (336, 1344), (336, 1680), (336, 2016), (672, 336), (672, 672), (672, 1008), (672, 1344), (672, 1680), (672, 2016), (1008, 336), (1008, 672), (1008, 1008), (1008, 1344), (1008, 1680), (1008, 2016), (1344, 336), (1344, 672), (1344, 1008), (1344, 1344), (1344, 1680), (1344, 2016), (1680, 336), (1680, 672), (1680, 1008), (1680, 1344), (1680, 1680), (1680, 2016), (2016, 336), (2016, 672), (2016, 1008), (2016, 1344), (2016, 1680), (2016, 2016)]", | |
| "image_seq_length": 576, | |
| "image_token_index": 200025, | |
| "video_token_index": 200026, | |
| "multimodal_projector_bias": true, | |
| "num_reward_heads": 0, | |
| "process_image_mode": "dynamic_res", | |
| "projector_hidden_act": "gelu", | |
| "vision_feature_layer": -1, | |
| "vision_feature_select_strategy": "full", | |
| "torch_dtype": "bfloat16", | |
| "transformers_version": "4.52.4", | |
| "projector_hidden_size": 6144, | |
| "quantization_config": { | |
| "global_quant_config": { | |
| "input_tensors": { | |
| "dtype": "fp4", | |
| "is_dynamic": true, | |
| "qscheme": "per_group", | |
| "ch_axis": -1, | |
| "group_size": 32, | |
| "block_size": null, | |
| "symmetric": null, | |
| "round_method": "half_even", | |
| "scale_type": "float", | |
| "scale_format": "e8m0", | |
| "scale_calculation_mode": "even", | |
| "mx_element_dtype": null, | |
| "observer_cls": "PerBlockMXObserver", | |
| "is_scale_quant": false, | |
| "enable_buffer_reuse": false, | |
| "max_input_numel": 4194304 | |
| }, | |
| "output_tensors": null, | |
| "weight": { | |
| "dtype": "fp4", | |
| "is_dynamic": false, | |
| "qscheme": "per_group", | |
| "ch_axis": -1, | |
| "group_size": 32, | |
| "block_size": null, | |
| "symmetric": null, | |
| "round_method": "half_even", | |
| "scale_type": "float", | |
| "scale_format": "e8m0", | |
| "scale_calculation_mode": "even", | |
| "mx_element_dtype": null, | |
| "observer_cls": "PerBlockMXObserver", | |
| "is_scale_quant": false, | |
| "enable_buffer_reuse": false, | |
| "max_input_numel": 4194304 | |
| }, | |
| "bias": null, | |
| "target_device": null | |
| }, | |
| "exclude": [ | |
| "language_model.lm_head", | |
| "language_model.model.layers.0.mlp.down_proj", | |
| "language_model.model.layers.0.mlp.gate_proj", | |
| "language_model.model.layers.0.mlp.up_proj", | |
| "language_model.model.layers.0.self_attn.k_proj", | |
| "language_model.model.layers.0.self_attn.o_proj", | |
| "language_model.model.layers.0.self_attn.q_proj", | |
| "language_model.model.layers.0.self_attn.v_proj", | |
| "language_model.model.layers.1.mlp.down_proj", | |
| "language_model.model.layers.1.mlp.gate_proj", | |
| "language_model.model.layers.1.mlp.up_proj", | |
| "language_model.model.layers.1.self_attn.k_proj", | |
| "language_model.model.layers.1.self_attn.o_proj", | |
| "language_model.model.layers.1.self_attn.q_proj", | |
| "language_model.model.layers.1.self_attn.v_proj", | |
| "language_model.model.layers.10.block_sparse_moe.gate", | |
| "language_model.model.layers.10.self_attn.index_k_proj", | |
| "language_model.model.layers.10.self_attn.index_q_proj", | |
| "language_model.model.layers.10.self_attn.k_proj", | |
| "language_model.model.layers.10.self_attn.o_proj", | |
| "language_model.model.layers.10.self_attn.q_proj", | |
| "language_model.model.layers.10.self_attn.v_proj", | |
| "language_model.model.layers.11.block_sparse_moe.gate", | |
| "language_model.model.layers.11.self_attn.index_k_proj", | |
| "language_model.model.layers.11.self_attn.index_q_proj", | |
| "language_model.model.layers.11.self_attn.k_proj", | |
| "language_model.model.layers.11.self_attn.o_proj", | |
| "language_model.model.layers.11.self_attn.q_proj", | |
| "language_model.model.layers.11.self_attn.v_proj", | |
| "language_model.model.layers.12.block_sparse_moe.gate", | |
| "language_model.model.layers.12.self_attn.index_k_proj", | |
| "language_model.model.layers.12.self_attn.index_q_proj", | |
| "language_model.model.layers.12.self_attn.k_proj", | |
| "language_model.model.layers.12.self_attn.o_proj", | |
| "language_model.model.layers.12.self_attn.q_proj", | |
| "language_model.model.layers.12.self_attn.v_proj", | |
| "language_model.model.layers.13.block_sparse_moe.gate", | |
| "language_model.model.layers.13.self_attn.index_k_proj", | |
| "language_model.model.layers.13.self_attn.index_q_proj", | |
| "language_model.model.layers.13.self_attn.k_proj", | |
| "language_model.model.layers.13.self_attn.o_proj", | |
| "language_model.model.layers.13.self_attn.q_proj", | |
| "language_model.model.layers.13.self_attn.v_proj", | |
| "language_model.model.layers.14.block_sparse_moe.gate", | |
| "language_model.model.layers.14.self_attn.index_k_proj", | |
| "language_model.model.layers.14.self_attn.index_q_proj", | |
| "language_model.model.layers.14.self_attn.k_proj", | |
| "language_model.model.layers.14.self_attn.o_proj", | |
| "language_model.model.layers.14.self_attn.q_proj", | |
| "language_model.model.layers.14.self_attn.v_proj", | |
| "language_model.model.layers.15.block_sparse_moe.gate", | |
| "language_model.model.layers.15.self_attn.index_k_proj", | |
| "language_model.model.layers.15.self_attn.index_q_proj", | |
| "language_model.model.layers.15.self_attn.k_proj", | |
| "language_model.model.layers.15.self_attn.o_proj", | |
| "language_model.model.layers.15.self_attn.q_proj", | |
| "language_model.model.layers.15.self_attn.v_proj", | |
| "language_model.model.layers.16.block_sparse_moe.gate", | |
| "language_model.model.layers.16.self_attn.index_k_proj", | |
| "language_model.model.layers.16.self_attn.index_q_proj", | |
| "language_model.model.layers.16.self_attn.k_proj", | |
| "language_model.model.layers.16.self_attn.o_proj", | |
| "language_model.model.layers.16.self_attn.q_proj", | |
| "language_model.model.layers.16.self_attn.v_proj", | |
| "language_model.model.layers.17.block_sparse_moe.gate", | |
| "language_model.model.layers.17.self_attn.index_k_proj", | |
| "language_model.model.layers.17.self_attn.index_q_proj", | |
| "language_model.model.layers.17.self_attn.k_proj", | |
| "language_model.model.layers.17.self_attn.o_proj", | |
| "language_model.model.layers.17.self_attn.q_proj", | |
| "language_model.model.layers.17.self_attn.v_proj", | |
| "language_model.model.layers.18.block_sparse_moe.gate", | |
| "language_model.model.layers.18.self_attn.index_k_proj", | |
| "language_model.model.layers.18.self_attn.index_q_proj", | |
| "language_model.model.layers.18.self_attn.k_proj", | |
| "language_model.model.layers.18.self_attn.o_proj", | |
| "language_model.model.layers.18.self_attn.q_proj", | |
| "language_model.model.layers.18.self_attn.v_proj", | |
| "language_model.model.layers.19.block_sparse_moe.gate", | |
| "language_model.model.layers.19.self_attn.index_k_proj", | |
| "language_model.model.layers.19.self_attn.index_q_proj", | |
| "language_model.model.layers.19.self_attn.k_proj", | |
| "language_model.model.layers.19.self_attn.o_proj", | |
| "language_model.model.layers.19.self_attn.q_proj", | |
| "language_model.model.layers.19.self_attn.v_proj", | |
| "language_model.model.layers.2.mlp.down_proj", | |
| "language_model.model.layers.2.mlp.gate_proj", | |
| "language_model.model.layers.2.mlp.up_proj", | |
| "language_model.model.layers.2.self_attn.k_proj", | |
| "language_model.model.layers.2.self_attn.o_proj", | |
| "language_model.model.layers.2.self_attn.q_proj", | |
| "language_model.model.layers.2.self_attn.v_proj", | |
| "language_model.model.layers.20.block_sparse_moe.gate", | |
| "language_model.model.layers.20.self_attn.index_k_proj", | |
| "language_model.model.layers.20.self_attn.index_q_proj", | |
| "language_model.model.layers.20.self_attn.k_proj", | |
| "language_model.model.layers.20.self_attn.o_proj", | |
| "language_model.model.layers.20.self_attn.q_proj", | |
| "language_model.model.layers.20.self_attn.v_proj", | |
| "language_model.model.layers.21.block_sparse_moe.gate", | |
| "language_model.model.layers.21.self_attn.index_k_proj", | |
| "language_model.model.layers.21.self_attn.index_q_proj", | |
| "language_model.model.layers.21.self_attn.k_proj", | |
| "language_model.model.layers.21.self_attn.o_proj", | |
| "language_model.model.layers.21.self_attn.q_proj", | |
| "language_model.model.layers.21.self_attn.v_proj", | |
| "language_model.model.layers.22.block_sparse_moe.gate", | |
| "language_model.model.layers.22.self_attn.index_k_proj", | |
| "language_model.model.layers.22.self_attn.index_q_proj", | |
| "language_model.model.layers.22.self_attn.k_proj", | |
| "language_model.model.layers.22.self_attn.o_proj", | |
| "language_model.model.layers.22.self_attn.q_proj", | |
| "language_model.model.layers.22.self_attn.v_proj", | |
| "language_model.model.layers.23.block_sparse_moe.gate", | |
| "language_model.model.layers.23.self_attn.index_k_proj", | |
| "language_model.model.layers.23.self_attn.index_q_proj", | |
| "language_model.model.layers.23.self_attn.k_proj", | |
| "language_model.model.layers.23.self_attn.o_proj", | |
| "language_model.model.layers.23.self_attn.q_proj", | |
| "language_model.model.layers.23.self_attn.v_proj", | |
| "language_model.model.layers.24.block_sparse_moe.gate", | |
| "language_model.model.layers.24.self_attn.index_k_proj", | |
| "language_model.model.layers.24.self_attn.index_q_proj", | |
| "language_model.model.layers.24.self_attn.k_proj", | |
| "language_model.model.layers.24.self_attn.o_proj", | |
| "language_model.model.layers.24.self_attn.q_proj", | |
| "language_model.model.layers.24.self_attn.v_proj", | |
| "language_model.model.layers.25.block_sparse_moe.gate", | |
| "language_model.model.layers.25.self_attn.index_k_proj", | |
| "language_model.model.layers.25.self_attn.index_q_proj", | |
| "language_model.model.layers.25.self_attn.k_proj", | |
| "language_model.model.layers.25.self_attn.o_proj", | |
| "language_model.model.layers.25.self_attn.q_proj", | |
| "language_model.model.layers.25.self_attn.v_proj", | |
| "language_model.model.layers.26.block_sparse_moe.gate", | |
| "language_model.model.layers.26.self_attn.index_k_proj", | |
| "language_model.model.layers.26.self_attn.index_q_proj", | |
| "language_model.model.layers.26.self_attn.k_proj", | |
| "language_model.model.layers.26.self_attn.o_proj", | |
| "language_model.model.layers.26.self_attn.q_proj", | |
| "language_model.model.layers.26.self_attn.v_proj", | |
| "language_model.model.layers.27.block_sparse_moe.gate", | |
| "language_model.model.layers.27.self_attn.index_k_proj", | |
| "language_model.model.layers.27.self_attn.index_q_proj", | |
| "language_model.model.layers.27.self_attn.k_proj", | |
| "language_model.model.layers.27.self_attn.o_proj", | |
| "language_model.model.layers.27.self_attn.q_proj", | |
| "language_model.model.layers.27.self_attn.v_proj", | |
| "language_model.model.layers.28.block_sparse_moe.gate", | |
| "language_model.model.layers.28.self_attn.index_k_proj", | |
| "language_model.model.layers.28.self_attn.index_q_proj", | |
| "language_model.model.layers.28.self_attn.k_proj", | |
| "language_model.model.layers.28.self_attn.o_proj", | |
| "language_model.model.layers.28.self_attn.q_proj", | |
| "language_model.model.layers.28.self_attn.v_proj", | |
| "language_model.model.layers.29.block_sparse_moe.gate", | |
| "language_model.model.layers.29.self_attn.index_k_proj", | |
| "language_model.model.layers.29.self_attn.index_q_proj", | |
| "language_model.model.layers.29.self_attn.k_proj", | |
| "language_model.model.layers.29.self_attn.o_proj", | |
| "language_model.model.layers.29.self_attn.q_proj", | |
| "language_model.model.layers.29.self_attn.v_proj", | |
| "language_model.model.layers.3.block_sparse_moe.gate", | |
| "language_model.model.layers.3.self_attn.index_k_proj", | |
| "language_model.model.layers.3.self_attn.index_q_proj", | |
| "language_model.model.layers.3.self_attn.k_proj", | |
| "language_model.model.layers.3.self_attn.o_proj", | |
| "language_model.model.layers.3.self_attn.q_proj", | |
| "language_model.model.layers.3.self_attn.v_proj", | |
| "language_model.model.layers.30.block_sparse_moe.gate", | |
| "language_model.model.layers.30.self_attn.index_k_proj", | |
| "language_model.model.layers.30.self_attn.index_q_proj", | |
| "language_model.model.layers.30.self_attn.k_proj", | |
| "language_model.model.layers.30.self_attn.o_proj", | |
| "language_model.model.layers.30.self_attn.q_proj", | |
| "language_model.model.layers.30.self_attn.v_proj", | |
| "language_model.model.layers.31.block_sparse_moe.gate", | |
| "language_model.model.layers.31.self_attn.index_k_proj", | |
| "language_model.model.layers.31.self_attn.index_q_proj", | |
| "language_model.model.layers.31.self_attn.k_proj", | |
| "language_model.model.layers.31.self_attn.o_proj", | |
| "language_model.model.layers.31.self_attn.q_proj", | |
| "language_model.model.layers.31.self_attn.v_proj", | |
| "language_model.model.layers.32.block_sparse_moe.gate", | |
| "language_model.model.layers.32.self_attn.index_k_proj", | |
| "language_model.model.layers.32.self_attn.index_q_proj", | |
| "language_model.model.layers.32.self_attn.k_proj", | |
| "language_model.model.layers.32.self_attn.o_proj", | |
| "language_model.model.layers.32.self_attn.q_proj", | |
| "language_model.model.layers.32.self_attn.v_proj", | |
| "language_model.model.layers.33.block_sparse_moe.gate", | |
| "language_model.model.layers.33.self_attn.index_k_proj", | |
| "language_model.model.layers.33.self_attn.index_q_proj", | |
| "language_model.model.layers.33.self_attn.k_proj", | |
| "language_model.model.layers.33.self_attn.o_proj", | |
| "language_model.model.layers.33.self_attn.q_proj", | |
| "language_model.model.layers.33.self_attn.v_proj", | |
| "language_model.model.layers.34.block_sparse_moe.gate", | |
| "language_model.model.layers.34.self_attn.index_k_proj", | |
| "language_model.model.layers.34.self_attn.index_q_proj", | |
| "language_model.model.layers.34.self_attn.k_proj", | |
| "language_model.model.layers.34.self_attn.o_proj", | |
| "language_model.model.layers.34.self_attn.q_proj", | |
| "language_model.model.layers.34.self_attn.v_proj", | |
| "language_model.model.layers.35.block_sparse_moe.gate", | |
| "language_model.model.layers.35.self_attn.index_k_proj", | |
| "language_model.model.layers.35.self_attn.index_q_proj", | |
| "language_model.model.layers.35.self_attn.k_proj", | |
| "language_model.model.layers.35.self_attn.o_proj", | |
| "language_model.model.layers.35.self_attn.q_proj", | |
| "language_model.model.layers.35.self_attn.v_proj", | |
| "language_model.model.layers.36.block_sparse_moe.gate", | |
| "language_model.model.layers.36.self_attn.index_k_proj", | |
| "language_model.model.layers.36.self_attn.index_q_proj", | |
| "language_model.model.layers.36.self_attn.k_proj", | |
| "language_model.model.layers.36.self_attn.o_proj", | |
| "language_model.model.layers.36.self_attn.q_proj", | |
| "language_model.model.layers.36.self_attn.v_proj", | |
| "language_model.model.layers.37.block_sparse_moe.gate", | |
| "language_model.model.layers.37.self_attn.index_k_proj", | |
| "language_model.model.layers.37.self_attn.index_q_proj", | |
| "language_model.model.layers.37.self_attn.k_proj", | |
| "language_model.model.layers.37.self_attn.o_proj", | |
| "language_model.model.layers.37.self_attn.q_proj", | |
| "language_model.model.layers.37.self_attn.v_proj", | |
| "language_model.model.layers.38.block_sparse_moe.gate", | |
| "language_model.model.layers.38.self_attn.index_k_proj", | |
| "language_model.model.layers.38.self_attn.index_q_proj", | |
| "language_model.model.layers.38.self_attn.k_proj", | |
| "language_model.model.layers.38.self_attn.o_proj", | |
| "language_model.model.layers.38.self_attn.q_proj", | |
| "language_model.model.layers.38.self_attn.v_proj", | |
| "language_model.model.layers.39.block_sparse_moe.gate", | |
| "language_model.model.layers.39.self_attn.index_k_proj", | |
| "language_model.model.layers.39.self_attn.index_q_proj", | |
| "language_model.model.layers.39.self_attn.k_proj", | |
| "language_model.model.layers.39.self_attn.o_proj", | |
| "language_model.model.layers.39.self_attn.q_proj", | |
| "language_model.model.layers.39.self_attn.v_proj", | |
| "language_model.model.layers.4.block_sparse_moe.gate", | |
| "language_model.model.layers.4.self_attn.index_k_proj", | |
| "language_model.model.layers.4.self_attn.index_q_proj", | |
| "language_model.model.layers.4.self_attn.k_proj", | |
| "language_model.model.layers.4.self_attn.o_proj", | |
| "language_model.model.layers.4.self_attn.q_proj", | |
| "language_model.model.layers.4.self_attn.v_proj", | |
| "language_model.model.layers.40.block_sparse_moe.gate", | |
| "language_model.model.layers.40.self_attn.index_k_proj", | |
| "language_model.model.layers.40.self_attn.index_q_proj", | |
| "language_model.model.layers.40.self_attn.k_proj", | |
| "language_model.model.layers.40.self_attn.o_proj", | |
| "language_model.model.layers.40.self_attn.q_proj", | |
| "language_model.model.layers.40.self_attn.v_proj", | |
| "language_model.model.layers.41.block_sparse_moe.gate", | |
| "language_model.model.layers.41.self_attn.index_k_proj", | |
| "language_model.model.layers.41.self_attn.index_q_proj", | |
| "language_model.model.layers.41.self_attn.k_proj", | |
| "language_model.model.layers.41.self_attn.o_proj", | |
| "language_model.model.layers.41.self_attn.q_proj", | |
| "language_model.model.layers.41.self_attn.v_proj", | |
| "language_model.model.layers.42.block_sparse_moe.gate", | |
| "language_model.model.layers.42.self_attn.index_k_proj", | |
| "language_model.model.layers.42.self_attn.index_q_proj", | |
| "language_model.model.layers.42.self_attn.k_proj", | |
| "language_model.model.layers.42.self_attn.o_proj", | |
| "language_model.model.layers.42.self_attn.q_proj", | |
| "language_model.model.layers.42.self_attn.v_proj", | |
| "language_model.model.layers.43.block_sparse_moe.gate", | |
| "language_model.model.layers.43.self_attn.index_k_proj", | |
| "language_model.model.layers.43.self_attn.index_q_proj", | |
| "language_model.model.layers.43.self_attn.k_proj", | |
| "language_model.model.layers.43.self_attn.o_proj", | |
| "language_model.model.layers.43.self_attn.q_proj", | |
| "language_model.model.layers.43.self_attn.v_proj", | |
| "language_model.model.layers.44.block_sparse_moe.gate", | |
| "language_model.model.layers.44.self_attn.index_k_proj", | |
| "language_model.model.layers.44.self_attn.index_q_proj", | |
| "language_model.model.layers.44.self_attn.k_proj", | |
| "language_model.model.layers.44.self_attn.o_proj", | |
| "language_model.model.layers.44.self_attn.q_proj", | |
| "language_model.model.layers.44.self_attn.v_proj", | |
| "language_model.model.layers.45.block_sparse_moe.gate", | |
| "language_model.model.layers.45.self_attn.index_k_proj", | |
| "language_model.model.layers.45.self_attn.index_q_proj", | |
| "language_model.model.layers.45.self_attn.k_proj", | |
| "language_model.model.layers.45.self_attn.o_proj", | |
| "language_model.model.layers.45.self_attn.q_proj", | |
| "language_model.model.layers.45.self_attn.v_proj", | |
| "language_model.model.layers.46.block_sparse_moe.gate", | |
| "language_model.model.layers.46.self_attn.index_k_proj", | |
| "language_model.model.layers.46.self_attn.index_q_proj", | |
| "language_model.model.layers.46.self_attn.k_proj", | |
| "language_model.model.layers.46.self_attn.o_proj", | |
| "language_model.model.layers.46.self_attn.q_proj", | |
| "language_model.model.layers.46.self_attn.v_proj", | |
| "language_model.model.layers.47.block_sparse_moe.gate", | |
| "language_model.model.layers.47.self_attn.index_k_proj", | |
| "language_model.model.layers.47.self_attn.index_q_proj", | |
| "language_model.model.layers.47.self_attn.k_proj", | |
| "language_model.model.layers.47.self_attn.o_proj", | |
| "language_model.model.layers.47.self_attn.q_proj", | |
| "language_model.model.layers.47.self_attn.v_proj", | |
| "language_model.model.layers.48.block_sparse_moe.gate", | |
| "language_model.model.layers.48.self_attn.index_k_proj", | |
| "language_model.model.layers.48.self_attn.index_q_proj", | |
| "language_model.model.layers.48.self_attn.k_proj", | |
| "language_model.model.layers.48.self_attn.o_proj", | |
| "language_model.model.layers.48.self_attn.q_proj", | |
| "language_model.model.layers.48.self_attn.v_proj", | |
| "language_model.model.layers.49.block_sparse_moe.gate", | |
| "language_model.model.layers.49.self_attn.index_k_proj", | |
| "language_model.model.layers.49.self_attn.index_q_proj", | |
| "language_model.model.layers.49.self_attn.k_proj", | |
| "language_model.model.layers.49.self_attn.o_proj", | |
| "language_model.model.layers.49.self_attn.q_proj", | |
| "language_model.model.layers.49.self_attn.v_proj", | |
| "language_model.model.layers.5.block_sparse_moe.gate", | |
| "language_model.model.layers.5.self_attn.index_k_proj", | |
| "language_model.model.layers.5.self_attn.index_q_proj", | |
| "language_model.model.layers.5.self_attn.k_proj", | |
| "language_model.model.layers.5.self_attn.o_proj", | |
| "language_model.model.layers.5.self_attn.q_proj", | |
| "language_model.model.layers.5.self_attn.v_proj", | |
| "language_model.model.layers.50.block_sparse_moe.gate", | |
| "language_model.model.layers.50.self_attn.index_k_proj", | |
| "language_model.model.layers.50.self_attn.index_q_proj", | |
| "language_model.model.layers.50.self_attn.k_proj", | |
| "language_model.model.layers.50.self_attn.o_proj", | |
| "language_model.model.layers.50.self_attn.q_proj", | |
| "language_model.model.layers.50.self_attn.v_proj", | |
| "language_model.model.layers.51.block_sparse_moe.gate", | |
| "language_model.model.layers.51.self_attn.index_k_proj", | |
| "language_model.model.layers.51.self_attn.index_q_proj", | |
| "language_model.model.layers.51.self_attn.k_proj", | |
| "language_model.model.layers.51.self_attn.o_proj", | |
| "language_model.model.layers.51.self_attn.q_proj", | |
| "language_model.model.layers.51.self_attn.v_proj", | |
| "language_model.model.layers.52.block_sparse_moe.gate", | |
| "language_model.model.layers.52.self_attn.index_k_proj", | |
| "language_model.model.layers.52.self_attn.index_q_proj", | |
| "language_model.model.layers.52.self_attn.k_proj", | |
| "language_model.model.layers.52.self_attn.o_proj", | |
| "language_model.model.layers.52.self_attn.q_proj", | |
| "language_model.model.layers.52.self_attn.v_proj", | |
| "language_model.model.layers.53.block_sparse_moe.gate", | |
| "language_model.model.layers.53.self_attn.index_k_proj", | |
| "language_model.model.layers.53.self_attn.index_q_proj", | |
| "language_model.model.layers.53.self_attn.k_proj", | |
| "language_model.model.layers.53.self_attn.o_proj", | |
| "language_model.model.layers.53.self_attn.q_proj", | |
| "language_model.model.layers.53.self_attn.v_proj", | |
| "language_model.model.layers.54.block_sparse_moe.gate", | |
| "language_model.model.layers.54.self_attn.index_k_proj", | |
| "language_model.model.layers.54.self_attn.index_q_proj", | |
| "language_model.model.layers.54.self_attn.k_proj", | |
| "language_model.model.layers.54.self_attn.o_proj", | |
| "language_model.model.layers.54.self_attn.q_proj", | |
| "language_model.model.layers.54.self_attn.v_proj", | |
| "language_model.model.layers.55.block_sparse_moe.gate", | |
| "language_model.model.layers.55.self_attn.index_k_proj", | |
| "language_model.model.layers.55.self_attn.index_q_proj", | |
| "language_model.model.layers.55.self_attn.k_proj", | |
| "language_model.model.layers.55.self_attn.o_proj", | |
| "language_model.model.layers.55.self_attn.q_proj", | |
| "language_model.model.layers.55.self_attn.v_proj", | |
| "language_model.model.layers.56.block_sparse_moe.gate", | |
| "language_model.model.layers.56.self_attn.index_k_proj", | |
| "language_model.model.layers.56.self_attn.index_q_proj", | |
| "language_model.model.layers.56.self_attn.k_proj", | |
| "language_model.model.layers.56.self_attn.o_proj", | |
| "language_model.model.layers.56.self_attn.q_proj", | |
| "language_model.model.layers.56.self_attn.v_proj", | |
| "language_model.model.layers.57.block_sparse_moe.gate", | |
| "language_model.model.layers.57.self_attn.index_k_proj", | |
| "language_model.model.layers.57.self_attn.index_q_proj", | |
| "language_model.model.layers.57.self_attn.k_proj", | |
| "language_model.model.layers.57.self_attn.o_proj", | |
| "language_model.model.layers.57.self_attn.q_proj", | |
| "language_model.model.layers.57.self_attn.v_proj", | |
| "language_model.model.layers.58.block_sparse_moe.gate", | |
| "language_model.model.layers.58.self_attn.index_k_proj", | |
| "language_model.model.layers.58.self_attn.index_q_proj", | |
| "language_model.model.layers.58.self_attn.k_proj", | |
| "language_model.model.layers.58.self_attn.o_proj", | |
| "language_model.model.layers.58.self_attn.q_proj", | |
| "language_model.model.layers.58.self_attn.v_proj", | |
| "language_model.model.layers.59.block_sparse_moe.gate", | |
| "language_model.model.layers.59.self_attn.index_k_proj", | |
| "language_model.model.layers.59.self_attn.index_q_proj", | |
| "language_model.model.layers.59.self_attn.k_proj", | |
| "language_model.model.layers.59.self_attn.o_proj", | |
| "language_model.model.layers.59.self_attn.q_proj", | |
| "language_model.model.layers.59.self_attn.v_proj", | |
| "language_model.model.layers.6.block_sparse_moe.gate", | |
| "language_model.model.layers.6.self_attn.index_k_proj", | |
| "language_model.model.layers.6.self_attn.index_q_proj", | |
| "language_model.model.layers.6.self_attn.k_proj", | |
| "language_model.model.layers.6.self_attn.o_proj", | |
| "language_model.model.layers.6.self_attn.q_proj", | |
| "language_model.model.layers.6.self_attn.v_proj", | |
| "language_model.model.layers.7.block_sparse_moe.gate", | |
| "language_model.model.layers.7.self_attn.index_k_proj", | |
| "language_model.model.layers.7.self_attn.index_q_proj", | |
| "language_model.model.layers.7.self_attn.k_proj", | |
| "language_model.model.layers.7.self_attn.o_proj", | |
| "language_model.model.layers.7.self_attn.q_proj", | |
| "language_model.model.layers.7.self_attn.v_proj", | |
| "language_model.model.layers.8.block_sparse_moe.gate", | |
| "language_model.model.layers.8.self_attn.index_k_proj", | |
| "language_model.model.layers.8.self_attn.index_q_proj", | |
| "language_model.model.layers.8.self_attn.k_proj", | |
| "language_model.model.layers.8.self_attn.o_proj", | |
| "language_model.model.layers.8.self_attn.q_proj", | |
| "language_model.model.layers.8.self_attn.v_proj", | |
| "language_model.model.layers.9.block_sparse_moe.gate", | |
| "language_model.model.layers.9.self_attn.index_k_proj", | |
| "language_model.model.layers.9.self_attn.index_q_proj", | |
| "language_model.model.layers.9.self_attn.k_proj", | |
| "language_model.model.layers.9.self_attn.o_proj", | |
| "language_model.model.layers.9.self_attn.q_proj", | |
| "language_model.model.layers.9.self_attn.v_proj", | |
| "multi_modal_projector.linear_1", | |
| "multi_modal_projector.linear_2", | |
| "patch_merge_mlp.linear_1", | |
| "patch_merge_mlp.linear_2", | |
| "vision_tower.vision_model.encoder.layers.0.layer_norm1", | |
| "vision_tower.vision_model.encoder.layers.0.layer_norm2", | |
| "vision_tower.vision_model.encoder.layers.0.mlp.fc1", | |
| "vision_tower.vision_model.encoder.layers.0.mlp.fc2", | |
| "vision_tower.vision_model.encoder.layers.0.self_attn.k_proj", | |
| "vision_tower.vision_model.encoder.layers.0.self_attn.out_proj", | |
| "vision_tower.vision_model.encoder.layers.0.self_attn.q_proj", | |
| "vision_tower.vision_model.encoder.layers.0.self_attn.v_proj", | |
| "vision_tower.vision_model.encoder.layers.1.layer_norm1", | |
| "vision_tower.vision_model.encoder.layers.1.layer_norm2", | |
| "vision_tower.vision_model.encoder.layers.1.mlp.fc1", | |
| "vision_tower.vision_model.encoder.layers.1.mlp.fc2", | |
| "vision_tower.vision_model.encoder.layers.1.self_attn.k_proj", | |
| "vision_tower.vision_model.encoder.layers.1.self_attn.out_proj", | |
| "vision_tower.vision_model.encoder.layers.1.self_attn.q_proj", | |
| "vision_tower.vision_model.encoder.layers.1.self_attn.v_proj", | |
| "vision_tower.vision_model.encoder.layers.10.layer_norm1", | |
| "vision_tower.vision_model.encoder.layers.10.layer_norm2", | |
| "vision_tower.vision_model.encoder.layers.10.mlp.fc1", | |
| "vision_tower.vision_model.encoder.layers.10.mlp.fc2", | |
| "vision_tower.vision_model.encoder.layers.10.self_attn.k_proj", | |
| "vision_tower.vision_model.encoder.layers.10.self_attn.out_proj", | |
| "vision_tower.vision_model.encoder.layers.10.self_attn.q_proj", | |
| "vision_tower.vision_model.encoder.layers.10.self_attn.v_proj", | |
| "vision_tower.vision_model.encoder.layers.11.layer_norm1", | |
| "vision_tower.vision_model.encoder.layers.11.layer_norm2", | |
| "vision_tower.vision_model.encoder.layers.11.mlp.fc1", | |
| "vision_tower.vision_model.encoder.layers.11.mlp.fc2", | |
| "vision_tower.vision_model.encoder.layers.11.self_attn.k_proj", | |
| "vision_tower.vision_model.encoder.layers.11.self_attn.out_proj", | |
| "vision_tower.vision_model.encoder.layers.11.self_attn.q_proj", | |
| "vision_tower.vision_model.encoder.layers.11.self_attn.v_proj", | |
| "vision_tower.vision_model.encoder.layers.12.layer_norm1", | |
| "vision_tower.vision_model.encoder.layers.12.layer_norm2", | |
| "vision_tower.vision_model.encoder.layers.12.mlp.fc1", | |
| "vision_tower.vision_model.encoder.layers.12.mlp.fc2", | |
| "vision_tower.vision_model.encoder.layers.12.self_attn.k_proj", | |
| "vision_tower.vision_model.encoder.layers.12.self_attn.out_proj", | |
| "vision_tower.vision_model.encoder.layers.12.self_attn.q_proj", | |
| "vision_tower.vision_model.encoder.layers.12.self_attn.v_proj", | |
| "vision_tower.vision_model.encoder.layers.13.layer_norm1", | |
| "vision_tower.vision_model.encoder.layers.13.layer_norm2", | |
| "vision_tower.vision_model.encoder.layers.13.mlp.fc1", | |
| "vision_tower.vision_model.encoder.layers.13.mlp.fc2", | |
| "vision_tower.vision_model.encoder.layers.13.self_attn.k_proj", | |
| "vision_tower.vision_model.encoder.layers.13.self_attn.out_proj", | |
| "vision_tower.vision_model.encoder.layers.13.self_attn.q_proj", | |
| "vision_tower.vision_model.encoder.layers.13.self_attn.v_proj", | |
| "vision_tower.vision_model.encoder.layers.14.layer_norm1", | |
| "vision_tower.vision_model.encoder.layers.14.layer_norm2", | |
| "vision_tower.vision_model.encoder.layers.14.mlp.fc1", | |
| "vision_tower.vision_model.encoder.layers.14.mlp.fc2", | |
| "vision_tower.vision_model.encoder.layers.14.self_attn.k_proj", | |
| "vision_tower.vision_model.encoder.layers.14.self_attn.out_proj", | |
| "vision_tower.vision_model.encoder.layers.14.self_attn.q_proj", | |
| "vision_tower.vision_model.encoder.layers.14.self_attn.v_proj", | |
| "vision_tower.vision_model.encoder.layers.15.layer_norm1", | |
| "vision_tower.vision_model.encoder.layers.15.layer_norm2", | |
| "vision_tower.vision_model.encoder.layers.15.mlp.fc1", | |
| "vision_tower.vision_model.encoder.layers.15.mlp.fc2", | |
| "vision_tower.vision_model.encoder.layers.15.self_attn.k_proj", | |
| "vision_tower.vision_model.encoder.layers.15.self_attn.out_proj", | |
| "vision_tower.vision_model.encoder.layers.15.self_attn.q_proj", | |
| "vision_tower.vision_model.encoder.layers.15.self_attn.v_proj", | |
| "vision_tower.vision_model.encoder.layers.16.layer_norm1", | |
| "vision_tower.vision_model.encoder.layers.16.layer_norm2", | |
| "vision_tower.vision_model.encoder.layers.16.mlp.fc1", | |
| "vision_tower.vision_model.encoder.layers.16.mlp.fc2", | |
| "vision_tower.vision_model.encoder.layers.16.self_attn.k_proj", | |
| "vision_tower.vision_model.encoder.layers.16.self_attn.out_proj", | |
| "vision_tower.vision_model.encoder.layers.16.self_attn.q_proj", | |
| "vision_tower.vision_model.encoder.layers.16.self_attn.v_proj", | |
| "vision_tower.vision_model.encoder.layers.17.layer_norm1", | |
| "vision_tower.vision_model.encoder.layers.17.layer_norm2", | |
| "vision_tower.vision_model.encoder.layers.17.mlp.fc1", | |
| "vision_tower.vision_model.encoder.layers.17.mlp.fc2", | |
| "vision_tower.vision_model.encoder.layers.17.self_attn.k_proj", | |
| "vision_tower.vision_model.encoder.layers.17.self_attn.out_proj", | |
| "vision_tower.vision_model.encoder.layers.17.self_attn.q_proj", | |
| "vision_tower.vision_model.encoder.layers.17.self_attn.v_proj", | |
| "vision_tower.vision_model.encoder.layers.18.layer_norm1", | |
| "vision_tower.vision_model.encoder.layers.18.layer_norm2", | |
| "vision_tower.vision_model.encoder.layers.18.mlp.fc1", | |
| "vision_tower.vision_model.encoder.layers.18.mlp.fc2", | |
| "vision_tower.vision_model.encoder.layers.18.self_attn.k_proj", | |
| "vision_tower.vision_model.encoder.layers.18.self_attn.out_proj", | |
| "vision_tower.vision_model.encoder.layers.18.self_attn.q_proj", | |
| "vision_tower.vision_model.encoder.layers.18.self_attn.v_proj", | |
| "vision_tower.vision_model.encoder.layers.19.layer_norm1", | |
| "vision_tower.vision_model.encoder.layers.19.layer_norm2", | |
| "vision_tower.vision_model.encoder.layers.19.mlp.fc1", | |
| "vision_tower.vision_model.encoder.layers.19.mlp.fc2", | |
| "vision_tower.vision_model.encoder.layers.19.self_attn.k_proj", | |
| "vision_tower.vision_model.encoder.layers.19.self_attn.out_proj", | |
| "vision_tower.vision_model.encoder.layers.19.self_attn.q_proj", | |
| "vision_tower.vision_model.encoder.layers.19.self_attn.v_proj", | |
| "vision_tower.vision_model.encoder.layers.2.layer_norm1", | |
| "vision_tower.vision_model.encoder.layers.2.layer_norm2", | |
| "vision_tower.vision_model.encoder.layers.2.mlp.fc1", | |
| "vision_tower.vision_model.encoder.layers.2.mlp.fc2", | |
| "vision_tower.vision_model.encoder.layers.2.self_attn.k_proj", | |
| "vision_tower.vision_model.encoder.layers.2.self_attn.out_proj", | |
| "vision_tower.vision_model.encoder.layers.2.self_attn.q_proj", | |
| "vision_tower.vision_model.encoder.layers.2.self_attn.v_proj", | |
| "vision_tower.vision_model.encoder.layers.20.layer_norm1", | |
| "vision_tower.vision_model.encoder.layers.20.layer_norm2", | |
| "vision_tower.vision_model.encoder.layers.20.mlp.fc1", | |
| "vision_tower.vision_model.encoder.layers.20.mlp.fc2", | |
| "vision_tower.vision_model.encoder.layers.20.self_attn.k_proj", | |
| "vision_tower.vision_model.encoder.layers.20.self_attn.out_proj", | |
| "vision_tower.vision_model.encoder.layers.20.self_attn.q_proj", | |
| "vision_tower.vision_model.encoder.layers.20.self_attn.v_proj", | |
| "vision_tower.vision_model.encoder.layers.21.layer_norm1", | |
| "vision_tower.vision_model.encoder.layers.21.layer_norm2", | |
| "vision_tower.vision_model.encoder.layers.21.mlp.fc1", | |
| "vision_tower.vision_model.encoder.layers.21.mlp.fc2", | |
| "vision_tower.vision_model.encoder.layers.21.self_attn.k_proj", | |
| "vision_tower.vision_model.encoder.layers.21.self_attn.out_proj", | |
| "vision_tower.vision_model.encoder.layers.21.self_attn.q_proj", | |
| "vision_tower.vision_model.encoder.layers.21.self_attn.v_proj", | |
| "vision_tower.vision_model.encoder.layers.22.layer_norm1", | |
| "vision_tower.vision_model.encoder.layers.22.layer_norm2", | |
| "vision_tower.vision_model.encoder.layers.22.mlp.fc1", | |
| "vision_tower.vision_model.encoder.layers.22.mlp.fc2", | |
| "vision_tower.vision_model.encoder.layers.22.self_attn.k_proj", | |
| "vision_tower.vision_model.encoder.layers.22.self_attn.out_proj", | |
| "vision_tower.vision_model.encoder.layers.22.self_attn.q_proj", | |
| "vision_tower.vision_model.encoder.layers.22.self_attn.v_proj", | |
| "vision_tower.vision_model.encoder.layers.23.layer_norm1", | |
| "vision_tower.vision_model.encoder.layers.23.layer_norm2", | |
| "vision_tower.vision_model.encoder.layers.23.mlp.fc1", | |
| "vision_tower.vision_model.encoder.layers.23.mlp.fc2", | |
| "vision_tower.vision_model.encoder.layers.23.self_attn.k_proj", | |
| "vision_tower.vision_model.encoder.layers.23.self_attn.out_proj", | |
| "vision_tower.vision_model.encoder.layers.23.self_attn.q_proj", | |
| "vision_tower.vision_model.encoder.layers.23.self_attn.v_proj", | |
| "vision_tower.vision_model.encoder.layers.24.layer_norm1", | |
| "vision_tower.vision_model.encoder.layers.24.layer_norm2", | |
| "vision_tower.vision_model.encoder.layers.24.mlp.fc1", | |
| "vision_tower.vision_model.encoder.layers.24.mlp.fc2", | |
| "vision_tower.vision_model.encoder.layers.24.self_attn.k_proj", | |
| "vision_tower.vision_model.encoder.layers.24.self_attn.out_proj", | |
| "vision_tower.vision_model.encoder.layers.24.self_attn.q_proj", | |
| "vision_tower.vision_model.encoder.layers.24.self_attn.v_proj", | |
| "vision_tower.vision_model.encoder.layers.25.layer_norm1", | |
| "vision_tower.vision_model.encoder.layers.25.layer_norm2", | |
| "vision_tower.vision_model.encoder.layers.25.mlp.fc1", | |
| "vision_tower.vision_model.encoder.layers.25.mlp.fc2", | |
| "vision_tower.vision_model.encoder.layers.25.self_attn.k_proj", | |
| "vision_tower.vision_model.encoder.layers.25.self_attn.out_proj", | |
| "vision_tower.vision_model.encoder.layers.25.self_attn.q_proj", | |
| "vision_tower.vision_model.encoder.layers.25.self_attn.v_proj", | |
| "vision_tower.vision_model.encoder.layers.26.layer_norm1", | |
| "vision_tower.vision_model.encoder.layers.26.layer_norm2", | |
| "vision_tower.vision_model.encoder.layers.26.mlp.fc1", | |
| "vision_tower.vision_model.encoder.layers.26.mlp.fc2", | |
| "vision_tower.vision_model.encoder.layers.26.self_attn.k_proj", | |
| "vision_tower.vision_model.encoder.layers.26.self_attn.out_proj", | |
| "vision_tower.vision_model.encoder.layers.26.self_attn.q_proj", | |
| "vision_tower.vision_model.encoder.layers.26.self_attn.v_proj", | |
| "vision_tower.vision_model.encoder.layers.27.layer_norm1", | |
| "vision_tower.vision_model.encoder.layers.27.layer_norm2", | |
| "vision_tower.vision_model.encoder.layers.27.mlp.fc1", | |
| "vision_tower.vision_model.encoder.layers.27.mlp.fc2", | |
| "vision_tower.vision_model.encoder.layers.27.self_attn.k_proj", | |
| "vision_tower.vision_model.encoder.layers.27.self_attn.out_proj", | |
| "vision_tower.vision_model.encoder.layers.27.self_attn.q_proj", | |
| "vision_tower.vision_model.encoder.layers.27.self_attn.v_proj", | |
| "vision_tower.vision_model.encoder.layers.28.layer_norm1", | |
| "vision_tower.vision_model.encoder.layers.28.layer_norm2", | |
| "vision_tower.vision_model.encoder.layers.28.mlp.fc1", | |
| "vision_tower.vision_model.encoder.layers.28.mlp.fc2", | |
| "vision_tower.vision_model.encoder.layers.28.self_attn.k_proj", | |
| "vision_tower.vision_model.encoder.layers.28.self_attn.out_proj", | |
| "vision_tower.vision_model.encoder.layers.28.self_attn.q_proj", | |
| "vision_tower.vision_model.encoder.layers.28.self_attn.v_proj", | |
| "vision_tower.vision_model.encoder.layers.29.layer_norm1", | |
| "vision_tower.vision_model.encoder.layers.29.layer_norm2", | |
| "vision_tower.vision_model.encoder.layers.29.mlp.fc1", | |
| "vision_tower.vision_model.encoder.layers.29.mlp.fc2", | |
| "vision_tower.vision_model.encoder.layers.29.self_attn.k_proj", | |
| "vision_tower.vision_model.encoder.layers.29.self_attn.out_proj", | |
| "vision_tower.vision_model.encoder.layers.29.self_attn.q_proj", | |
| "vision_tower.vision_model.encoder.layers.29.self_attn.v_proj", | |
| "vision_tower.vision_model.encoder.layers.3.layer_norm1", | |
| "vision_tower.vision_model.encoder.layers.3.layer_norm2", | |
| "vision_tower.vision_model.encoder.layers.3.mlp.fc1", | |
| "vision_tower.vision_model.encoder.layers.3.mlp.fc2", | |
| "vision_tower.vision_model.encoder.layers.3.self_attn.k_proj", | |
| "vision_tower.vision_model.encoder.layers.3.self_attn.out_proj", | |
| "vision_tower.vision_model.encoder.layers.3.self_attn.q_proj", | |
| "vision_tower.vision_model.encoder.layers.3.self_attn.v_proj", | |
| "vision_tower.vision_model.encoder.layers.30.layer_norm1", | |
| "vision_tower.vision_model.encoder.layers.30.layer_norm2", | |
| "vision_tower.vision_model.encoder.layers.30.mlp.fc1", | |
| "vision_tower.vision_model.encoder.layers.30.mlp.fc2", | |
| "vision_tower.vision_model.encoder.layers.30.self_attn.k_proj", | |
| "vision_tower.vision_model.encoder.layers.30.self_attn.out_proj", | |
| "vision_tower.vision_model.encoder.layers.30.self_attn.q_proj", | |
| "vision_tower.vision_model.encoder.layers.30.self_attn.v_proj", | |
| "vision_tower.vision_model.encoder.layers.31.layer_norm1", | |
| "vision_tower.vision_model.encoder.layers.31.layer_norm2", | |
| "vision_tower.vision_model.encoder.layers.31.mlp.fc1", | |
| "vision_tower.vision_model.encoder.layers.31.mlp.fc2", | |
| "vision_tower.vision_model.encoder.layers.31.self_attn.k_proj", | |
| "vision_tower.vision_model.encoder.layers.31.self_attn.out_proj", | |
| "vision_tower.vision_model.encoder.layers.31.self_attn.q_proj", | |
| "vision_tower.vision_model.encoder.layers.31.self_attn.v_proj", | |
| "vision_tower.vision_model.encoder.layers.4.layer_norm1", | |
| "vision_tower.vision_model.encoder.layers.4.layer_norm2", | |
| "vision_tower.vision_model.encoder.layers.4.mlp.fc1", | |
| "vision_tower.vision_model.encoder.layers.4.mlp.fc2", | |
| "vision_tower.vision_model.encoder.layers.4.self_attn.k_proj", | |
| "vision_tower.vision_model.encoder.layers.4.self_attn.out_proj", | |
| "vision_tower.vision_model.encoder.layers.4.self_attn.q_proj", | |
| "vision_tower.vision_model.encoder.layers.4.self_attn.v_proj", | |
| "vision_tower.vision_model.encoder.layers.5.layer_norm1", | |
| "vision_tower.vision_model.encoder.layers.5.layer_norm2", | |
| "vision_tower.vision_model.encoder.layers.5.mlp.fc1", | |
| "vision_tower.vision_model.encoder.layers.5.mlp.fc2", | |
| "vision_tower.vision_model.encoder.layers.5.self_attn.k_proj", | |
| "vision_tower.vision_model.encoder.layers.5.self_attn.out_proj", | |
| "vision_tower.vision_model.encoder.layers.5.self_attn.q_proj", | |
| "vision_tower.vision_model.encoder.layers.5.self_attn.v_proj", | |
| "vision_tower.vision_model.encoder.layers.6.layer_norm1", | |
| "vision_tower.vision_model.encoder.layers.6.layer_norm2", | |
| "vision_tower.vision_model.encoder.layers.6.mlp.fc1", | |
| "vision_tower.vision_model.encoder.layers.6.mlp.fc2", | |
| "vision_tower.vision_model.encoder.layers.6.self_attn.k_proj", | |
| "vision_tower.vision_model.encoder.layers.6.self_attn.out_proj", | |
| "vision_tower.vision_model.encoder.layers.6.self_attn.q_proj", | |
| "vision_tower.vision_model.encoder.layers.6.self_attn.v_proj", | |
| "vision_tower.vision_model.encoder.layers.7.layer_norm1", | |
| "vision_tower.vision_model.encoder.layers.7.layer_norm2", | |
| "vision_tower.vision_model.encoder.layers.7.mlp.fc1", | |
| "vision_tower.vision_model.encoder.layers.7.mlp.fc2", | |
| "vision_tower.vision_model.encoder.layers.7.self_attn.k_proj", | |
| "vision_tower.vision_model.encoder.layers.7.self_attn.out_proj", | |
| "vision_tower.vision_model.encoder.layers.7.self_attn.q_proj", | |
| "vision_tower.vision_model.encoder.layers.7.self_attn.v_proj", | |
| "vision_tower.vision_model.encoder.layers.8.layer_norm1", | |
| "vision_tower.vision_model.encoder.layers.8.layer_norm2", | |
| "vision_tower.vision_model.encoder.layers.8.mlp.fc1", | |
| "vision_tower.vision_model.encoder.layers.8.mlp.fc2", | |
| "vision_tower.vision_model.encoder.layers.8.self_attn.k_proj", | |
| "vision_tower.vision_model.encoder.layers.8.self_attn.out_proj", | |
| "vision_tower.vision_model.encoder.layers.8.self_attn.q_proj", | |
| "vision_tower.vision_model.encoder.layers.8.self_attn.v_proj", | |
| "vision_tower.vision_model.encoder.layers.9.layer_norm1", | |
| "vision_tower.vision_model.encoder.layers.9.layer_norm2", | |
| "vision_tower.vision_model.encoder.layers.9.mlp.fc1", | |
| "vision_tower.vision_model.encoder.layers.9.mlp.fc2", | |
| "vision_tower.vision_model.encoder.layers.9.self_attn.k_proj", | |
| "vision_tower.vision_model.encoder.layers.9.self_attn.out_proj", | |
| "vision_tower.vision_model.encoder.layers.9.self_attn.q_proj", | |
| "vision_tower.vision_model.encoder.layers.9.self_attn.v_proj" | |
| ], | |
| "algo_config": null, | |
| "softmax_quant_spec": null, | |
| "quant_method": "quark", | |
| "layer_type_quant_config": {}, | |
| "layer_quant_config": {}, | |
| "kv_cache_quant_config": {}, | |
| "kv_cache_post_rope": false, | |
| "quant_mode": "eager_mode", | |
| "version": "0.12+unknown", | |
| "export": { | |
| "kv_cache_group": [], | |
| "min_kv_scale": 0.0, | |
| "pack_method": "reorder", | |
| "weight_format": "real_quantized", | |
| "weight_merge_groups": null | |
| } | |
| } | |
| } |