Instructions to use zai-org/GLM-5.3-Flash with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use zai-org/GLM-5.3-Flash with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="zai-org/GLM-5.3-Flash")
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("zai-org/GLM-5.3-Flash")
model = AutoModelForMultimodalLM.from_pretrained("zai-org/GLM-5.3-Flash", 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]:]))

Inference
HuggingChat
Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use zai-org/GLM-5.3-Flash with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "zai-org/GLM-5.3-Flash"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "zai-org/GLM-5.3-Flash",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/zai-org/GLM-5.3-Flash

SGLang

How to use zai-org/GLM-5.3-Flash 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 "zai-org/GLM-5.3-Flash" \
    --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": "zai-org/GLM-5.3-Flash",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "zai-org/GLM-5.3-Flash" \
        --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": "zai-org/GLM-5.3-Flash",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use zai-org/GLM-5.3-Flash with Docker Model Runner:
```
docker model run hf.co/zai-org/GLM-5.3-Flash
```

GLM-5.3-Flash / config.json

zRzRzRzRzRzRzR

update

c54b8d1 about 6 hours ago

Raw

History Blame Contribute Delete

69.4 kB

	{
	"architectures": [
	"Glm5NextForConditionalGeneration"
	],
	"text_config": {
	"attention_bias": false,
	"attention_dropout": 0.0,
	"dtype": "bfloat16",
	"eos_token_id": [
	154820,
	154827,
	154829
	],
	"first_k_dense_replace": 3,
	"hc_eps": 1e-06,
	"hc_mult": 4,
	"hc_sinkhorn_iters": 20,
	"head_dim": 0,
	"hidden_act": "silu",
	"hidden_size": 4096,
	"index_head_dim": 128,
	"index_kpool": 4,
	"index_kpool_always_select_tail": true,
	"index_kpool_compress": true,
	"index_n_heads": 32,
	"index_topk": 2048,
	"index_share_for_mtp_iteration": true,
	"indexer_rope_interleave": true,
	"indexer_types": [
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full",
	"full"
	],
	"initializer_range": 0.02,
	"intermediate_size": 12288,
	"kv_lora_rank": 512,
	"layer_types": [
	"linear_attention",
	"linear_attention",
	"linear_attention",
	"deepseek_sparse_attention",
	"linear_attention",
	"linear_attention",
	"linear_attention",
	"deepseek_sparse_attention",
	"linear_attention",
	"linear_attention",
	"linear_attention",
	"deepseek_sparse_attention",
	"linear_attention",
	"linear_attention",
	"linear_attention",
	"deepseek_sparse_attention",
	"linear_attention",
	"linear_attention",
	"linear_attention",
	"deepseek_sparse_attention",
	"linear_attention",
	"linear_attention",
	"linear_attention",
	"deepseek_sparse_attention",
	"linear_attention",
	"linear_attention",
	"linear_attention",
	"deepseek_sparse_attention",
	"linear_attention",
	"linear_attention",
	"linear_attention",
	"deepseek_sparse_attention",
	"linear_attention",
	"linear_attention",
	"linear_attention",
	"deepseek_sparse_attention",
	"linear_attention",
	"linear_attention",
	"linear_attention",
	"deepseek_sparse_attention",
	"linear_attention",
	"linear_attention",
	"linear_attention",
	"deepseek_sparse_attention",
	"linear_attention"
	],
	"linear_attn_config": {
	"num_heads": 64,
	"gate_lower_bound": -5.0,
	"head_dim": 128,
	"short_conv_kernel_size": 4,
	"kda_layers": [
	0,
	1,
	2,
	4,
	5,
	6,
	8,
	9,
	10,
	12,
	13,
	14,
	16,
	17,
	18,
	20,
	21,
	22,
	24,
	25,
	26,
	28,
	29,
	30,
	32,
	33,
	34,
	36,
	37,
	38,
	40,
	41,
	42,
	44
	],
	"full_attn_layers": [
	3,
	7,
	11,
	15,
	19,
	23,
	27,
	31,
	35,
	39,
	43
	]
	},
	"max_position_embeddings": 1048576,
	"mhc": true,
	"mla_use_nope": true,
	"mlp_layer_types": [
	"dense",
	"dense",
	"dense",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse",
	"sparse"
	],
	"model_type": "glm5_next_text",
	"moe_intermediate_size": 2048,
	"moe_router_dtype": "float32",
	"n_group": 1,
	"n_routed_experts": 288,
	"n_shared_experts": 1,
	"norm_topk_prob": true,
	"num_attention_heads": 64,
	"num_experts_per_tok": 8,
	"num_hidden_layers": 45,
	"num_key_value_heads": 64,
	"num_nextn_predict_layers": 1,
	"output_router_logits": false,
	"pad_token_id": 154820,
	"q_lora_rank": 1536,
	"qk_head_dim": 256,
	"qk_nope_head_dim": 256,
	"qk_rope_head_dim": 0,
	"rms_norm_eps": 1e-05,
	"routed_scaling_factor": 2.5,
	"router_aux_loss_coef": 0.001,
	"scoring_func": "sigmoid",
	"swiglu_limit": 10.0,
	"tie_word_embeddings": false,
	"topk_group": 1,
	"topk_method": "noaux_tc",
	"use_cache": true,
	"v_head_dim": 256,
	"vocab_size": 154880
	},
	"vision_config": {
	"attention_bias": true,
	"attention_dropout": 0.0,
	"depth": 24,
	"hidden_act": "silu",
	"hidden_size": 1024,
	"image_size": 448,
	"in_channels": 3,
	"initializer_range": 0.02,
	"intermediate_size": 4096,
	"model_type": "glm5_next_vision",
	"num_heads": 16,
	"out_hidden_size": 4096,
	"patch_size": 14,
	"projection_intermediate_size": 10240,
	"rms_norm_eps": 1e-05,
	"spatial_merge_size": 2,
	"swiglu_limit": 10.0,
	"temporal_patch_size": 2
	},
	"image_token_id": 154854,
	"video_token_id": 154855,
	"image_start_token_id": 154830,
	"image_end_token_id": 154831,
	"video_start_token_id": 154832,
	"video_end_token_id": 154833,
	"tie_word_embeddings": false,
	"model_type": "glm5_next",
	"transformers_version": "5.16.0",
	"quantization_config": {
	"activation_scheme": "dynamic",
	"fmt": "e4m3",
	"modules_to_not_convert": [
	"attn_mha",
	"attn_mqa",
	"dt_bias",
	"hyper_connection",
	"lm_head",
	"mapping_proj",
	"model.embed_tokens",
	"model.layers.0.hc_attn_base",
	"model.layers.0.hc_attn_fn",
	"model.layers.0.hc_attn_scale",
	"model.layers.0.hc_ffn_base",
	"model.layers.0.hc_ffn_fn",
	"model.layers.0.hc_ffn_scale",
	"model.layers.0.input_layernorm",
	"model.layers.0.post_attention_layernorm",
	"model.layers.0.self_attn.A_log",
	"model.layers.0.self_attn.b_proj",
	"model.layers.0.self_attn.dt_bias",
	"model.layers.0.self_attn.f_a_proj",
	"model.layers.0.self_attn.f_b_proj",
	"model.layers.0.self_attn.fused_qkvbfg_a_proj",
	"model.layers.0.self_attn.g_a_proj",
	"model.layers.0.self_attn.g_b_proj",
	"model.layers.0.self_attn.k_conv1d",
	"model.layers.0.self_attn.k_proj",
	"model.layers.0.self_attn.o_norm",
	"model.layers.0.self_attn.o_proj",
	"model.layers.0.self_attn.q_conv1d",
	"model.layers.0.self_attn.q_proj",
	"model.layers.0.self_attn.qkv_proj",
	"model.layers.0.self_attn.v_conv1d",
	"model.layers.0.self_attn.v_proj",
	"model.layers.1.hc_attn_base",
	"model.layers.1.hc_attn_fn",
	"model.layers.1.hc_attn_scale",
	"model.layers.1.hc_ffn_base",
	"model.layers.1.hc_ffn_fn",
	"model.layers.1.hc_ffn_scale",
	"model.layers.1.input_layernorm",
	"model.layers.1.post_attention_layernorm",
	"model.layers.1.self_attn.A_log",
	"model.layers.1.self_attn.b_proj",
	"model.layers.1.self_attn.dt_bias",
	"model.layers.1.self_attn.f_a_proj",
	"model.layers.1.self_attn.f_b_proj",
	"model.layers.1.self_attn.fused_qkvbfg_a_proj",
	"model.layers.1.self_attn.g_a_proj",
	"model.layers.1.self_attn.g_b_proj",
	"model.layers.1.self_attn.k_conv1d",
	"model.layers.1.self_attn.k_proj",
	"model.layers.1.self_attn.o_norm",
	"model.layers.1.self_attn.o_proj",
	"model.layers.1.self_attn.q_conv1d",
	"model.layers.1.self_attn.q_proj",
	"model.layers.1.self_attn.qkv_proj",
	"model.layers.1.self_attn.v_conv1d",
	"model.layers.1.self_attn.v_proj",
	"model.layers.10.hc_attn_base",
	"model.layers.10.hc_attn_fn",
	"model.layers.10.hc_attn_scale",
	"model.layers.10.hc_ffn_base",
	"model.layers.10.hc_ffn_fn",
	"model.layers.10.hc_ffn_scale",
	"model.layers.10.input_layernorm",
	"model.layers.10.mlp.gate",
	"model.layers.10.mlp.gate.e_score_correction_bias",
	"model.layers.10.post_attention_layernorm",
	"model.layers.10.self_attn.A_log",
	"model.layers.10.self_attn.b_proj",
	"model.layers.10.self_attn.dt_bias",
	"model.layers.10.self_attn.f_a_proj",
	"model.layers.10.self_attn.f_b_proj",
	"model.layers.10.self_attn.fused_qkvbfg_a_proj",
	"model.layers.10.self_attn.g_a_proj",
	"model.layers.10.self_attn.g_b_proj",
	"model.layers.10.self_attn.k_conv1d",
	"model.layers.10.self_attn.k_proj",
	"model.layers.10.self_attn.o_norm",
	"model.layers.10.self_attn.o_proj",
	"model.layers.10.self_attn.q_conv1d",
	"model.layers.10.self_attn.q_proj",
	"model.layers.10.self_attn.qkv_proj",
	"model.layers.10.self_attn.v_conv1d",
	"model.layers.10.self_attn.v_proj",
	"model.layers.11.hc_attn_base",
	"model.layers.11.hc_attn_fn",
	"model.layers.11.hc_attn_scale",
	"model.layers.11.hc_ffn_base",
	"model.layers.11.hc_ffn_fn",
	"model.layers.11.hc_ffn_scale",
	"model.layers.11.input_layernorm",
	"model.layers.11.mlp.gate",
	"model.layers.11.mlp.gate.e_score_correction_bias",
	"model.layers.11.post_attention_layernorm",
	"model.layers.11.self_attn.indexer.index_kpool_compress_ape",
	"model.layers.11.self_attn.indexer.index_kpool_compress_gate",
	"model.layers.11.self_attn.indexer.k_norm",
	"model.layers.11.self_attn.indexer.k_norm.bias",
	"model.layers.11.self_attn.indexer.weights_proj",
	"model.layers.11.self_attn.indexer.wk",
	"model.layers.11.self_attn.indexer.wq_b",
	"model.layers.11.self_attn.kv_a_layernorm",
	"model.layers.11.self_attn.kv_b_proj",
	"model.layers.11.self_attn.q_a_layernorm",
	"model.layers.12.hc_attn_base",
	"model.layers.12.hc_attn_fn",
	"model.layers.12.hc_attn_scale",
	"model.layers.12.hc_ffn_base",
	"model.layers.12.hc_ffn_fn",
	"model.layers.12.hc_ffn_scale",
	"model.layers.12.input_layernorm",
	"model.layers.12.mlp.gate",
	"model.layers.12.mlp.gate.e_score_correction_bias",
	"model.layers.12.post_attention_layernorm",
	"model.layers.12.self_attn.A_log",
	"model.layers.12.self_attn.b_proj",
	"model.layers.12.self_attn.dt_bias",
	"model.layers.12.self_attn.f_a_proj",
	"model.layers.12.self_attn.f_b_proj",
	"model.layers.12.self_attn.fused_qkvbfg_a_proj",
	"model.layers.12.self_attn.g_a_proj",
	"model.layers.12.self_attn.g_b_proj",
	"model.layers.12.self_attn.k_conv1d",
	"model.layers.12.self_attn.k_proj",
	"model.layers.12.self_attn.o_norm",
	"model.layers.12.self_attn.o_proj",
	"model.layers.12.self_attn.q_conv1d",
	"model.layers.12.self_attn.q_proj",
	"model.layers.12.self_attn.qkv_proj",
	"model.layers.12.self_attn.v_conv1d",
	"model.layers.12.self_attn.v_proj",
	"model.layers.13.hc_attn_base",
	"model.layers.13.hc_attn_fn",
	"model.layers.13.hc_attn_scale",
	"model.layers.13.hc_ffn_base",
	"model.layers.13.hc_ffn_fn",
	"model.layers.13.hc_ffn_scale",
	"model.layers.13.input_layernorm",
	"model.layers.13.mlp.gate",
	"model.layers.13.mlp.gate.e_score_correction_bias",
	"model.layers.13.post_attention_layernorm",
	"model.layers.13.self_attn.A_log",
	"model.layers.13.self_attn.b_proj",
	"model.layers.13.self_attn.dt_bias",
	"model.layers.13.self_attn.f_a_proj",
	"model.layers.13.self_attn.f_b_proj",
	"model.layers.13.self_attn.fused_qkvbfg_a_proj",
	"model.layers.13.self_attn.g_a_proj",
	"model.layers.13.self_attn.g_b_proj",
	"model.layers.13.self_attn.k_conv1d",
	"model.layers.13.self_attn.k_proj",
	"model.layers.13.self_attn.o_norm",
	"model.layers.13.self_attn.o_proj",
	"model.layers.13.self_attn.q_conv1d",
	"model.layers.13.self_attn.q_proj",
	"model.layers.13.self_attn.qkv_proj",
	"model.layers.13.self_attn.v_conv1d",
	"model.layers.13.self_attn.v_proj",
	"model.layers.14.hc_attn_base",
	"model.layers.14.hc_attn_fn",
	"model.layers.14.hc_attn_scale",
	"model.layers.14.hc_ffn_base",
	"model.layers.14.hc_ffn_fn",
	"model.layers.14.hc_ffn_scale",
	"model.layers.14.input_layernorm",
	"model.layers.14.mlp.gate",
	"model.layers.14.mlp.gate.e_score_correction_bias",
	"model.layers.14.post_attention_layernorm",
	"model.layers.14.self_attn.A_log",
	"model.layers.14.self_attn.b_proj",
	"model.layers.14.self_attn.dt_bias",
	"model.layers.14.self_attn.f_a_proj",
	"model.layers.14.self_attn.f_b_proj",
	"model.layers.14.self_attn.fused_qkvbfg_a_proj",
	"model.layers.14.self_attn.g_a_proj",
	"model.layers.14.self_attn.g_b_proj",
	"model.layers.14.self_attn.k_conv1d",
	"model.layers.14.self_attn.k_proj",
	"model.layers.14.self_attn.o_norm",
	"model.layers.14.self_attn.o_proj",
	"model.layers.14.self_attn.q_conv1d",
	"model.layers.14.self_attn.q_proj",
	"model.layers.14.self_attn.qkv_proj",
	"model.layers.14.self_attn.v_conv1d",
	"model.layers.14.self_attn.v_proj",
	"model.layers.15.hc_attn_base",
	"model.layers.15.hc_attn_fn",
	"model.layers.15.hc_attn_scale",
	"model.layers.15.hc_ffn_base",
	"model.layers.15.hc_ffn_fn",
	"model.layers.15.hc_ffn_scale",
	"model.layers.15.input_layernorm",
	"model.layers.15.mlp.gate",
	"model.layers.15.mlp.gate.e_score_correction_bias",
	"model.layers.15.post_attention_layernorm",
	"model.layers.15.self_attn.indexer.index_kpool_compress_ape",
	"model.layers.15.self_attn.indexer.index_kpool_compress_gate",
	"model.layers.15.self_attn.indexer.k_norm",
	"model.layers.15.self_attn.indexer.k_norm.bias",
	"model.layers.15.self_attn.indexer.weights_proj",
	"model.layers.15.self_attn.indexer.wk",
	"model.layers.15.self_attn.indexer.wq_b",
	"model.layers.15.self_attn.kv_a_layernorm",
	"model.layers.15.self_attn.kv_b_proj",
	"model.layers.15.self_attn.q_a_layernorm",
	"model.layers.16.hc_attn_base",
	"model.layers.16.hc_attn_fn",
	"model.layers.16.hc_attn_scale",
	"model.layers.16.hc_ffn_base",
	"model.layers.16.hc_ffn_fn",
	"model.layers.16.hc_ffn_scale",
	"model.layers.16.input_layernorm",
	"model.layers.16.mlp.gate",
	"model.layers.16.mlp.gate.e_score_correction_bias",
	"model.layers.16.post_attention_layernorm",
	"model.layers.16.self_attn.A_log",
	"model.layers.16.self_attn.b_proj",
	"model.layers.16.self_attn.dt_bias",
	"model.layers.16.self_attn.f_a_proj",
	"model.layers.16.self_attn.f_b_proj",
	"model.layers.16.self_attn.fused_qkvbfg_a_proj",
	"model.layers.16.self_attn.g_a_proj",
	"model.layers.16.self_attn.g_b_proj",
	"model.layers.16.self_attn.k_conv1d",
	"model.layers.16.self_attn.k_proj",
	"model.layers.16.self_attn.o_norm",
	"model.layers.16.self_attn.o_proj",
	"model.layers.16.self_attn.q_conv1d",
	"model.layers.16.self_attn.q_proj",
	"model.layers.16.self_attn.qkv_proj",
	"model.layers.16.self_attn.v_conv1d",
	"model.layers.16.self_attn.v_proj",
	"model.layers.17.hc_attn_base",
	"model.layers.17.hc_attn_fn",
	"model.layers.17.hc_attn_scale",
	"model.layers.17.hc_ffn_base",
	"model.layers.17.hc_ffn_fn",
	"model.layers.17.hc_ffn_scale",
	"model.layers.17.input_layernorm",
	"model.layers.17.mlp.gate",
	"model.layers.17.mlp.gate.e_score_correction_bias",
	"model.layers.17.post_attention_layernorm",
	"model.layers.17.self_attn.A_log",
	"model.layers.17.self_attn.b_proj",
	"model.layers.17.self_attn.dt_bias",
	"model.layers.17.self_attn.f_a_proj",
	"model.layers.17.self_attn.f_b_proj",
	"model.layers.17.self_attn.fused_qkvbfg_a_proj",
	"model.layers.17.self_attn.g_a_proj",
	"model.layers.17.self_attn.g_b_proj",
	"model.layers.17.self_attn.k_conv1d",
	"model.layers.17.self_attn.k_proj",
	"model.layers.17.self_attn.o_norm",
	"model.layers.17.self_attn.o_proj",
	"model.layers.17.self_attn.q_conv1d",
	"model.layers.17.self_attn.q_proj",
	"model.layers.17.self_attn.qkv_proj",
	"model.layers.17.self_attn.v_conv1d",
	"model.layers.17.self_attn.v_proj",
	"model.layers.18.hc_attn_base",
	"model.layers.18.hc_attn_fn",
	"model.layers.18.hc_attn_scale",
	"model.layers.18.hc_ffn_base",
	"model.layers.18.hc_ffn_fn",
	"model.layers.18.hc_ffn_scale",
	"model.layers.18.input_layernorm",
	"model.layers.18.mlp.gate",
	"model.layers.18.mlp.gate.e_score_correction_bias",
	"model.layers.18.post_attention_layernorm",
	"model.layers.18.self_attn.A_log",
	"model.layers.18.self_attn.b_proj",
	"model.layers.18.self_attn.dt_bias",
	"model.layers.18.self_attn.f_a_proj",
	"model.layers.18.self_attn.f_b_proj",
	"model.layers.18.self_attn.fused_qkvbfg_a_proj",
	"model.layers.18.self_attn.g_a_proj",
	"model.layers.18.self_attn.g_b_proj",
	"model.layers.18.self_attn.k_conv1d",
	"model.layers.18.self_attn.k_proj",
	"model.layers.18.self_attn.o_norm",
	"model.layers.18.self_attn.o_proj",
	"model.layers.18.self_attn.q_conv1d",
	"model.layers.18.self_attn.q_proj",
	"model.layers.18.self_attn.qkv_proj",
	"model.layers.18.self_attn.v_conv1d",
	"model.layers.18.self_attn.v_proj",
	"model.layers.19.hc_attn_base",
	"model.layers.19.hc_attn_fn",
	"model.layers.19.hc_attn_scale",
	"model.layers.19.hc_ffn_base",
	"model.layers.19.hc_ffn_fn",
	"model.layers.19.hc_ffn_scale",
	"model.layers.19.input_layernorm",
	"model.layers.19.mlp.gate",
	"model.layers.19.mlp.gate.e_score_correction_bias",
	"model.layers.19.post_attention_layernorm",
	"model.layers.19.self_attn.indexer.index_kpool_compress_ape",
	"model.layers.19.self_attn.indexer.index_kpool_compress_gate",
	"model.layers.19.self_attn.indexer.k_norm",
	"model.layers.19.self_attn.indexer.k_norm.bias",
	"model.layers.19.self_attn.indexer.weights_proj",
	"model.layers.19.self_attn.indexer.wk",
	"model.layers.19.self_attn.indexer.wq_b",
	"model.layers.19.self_attn.kv_a_layernorm",
	"model.layers.19.self_attn.kv_b_proj",
	"model.layers.19.self_attn.q_a_layernorm",
	"model.layers.2.hc_attn_base",
	"model.layers.2.hc_attn_fn",
	"model.layers.2.hc_attn_scale",
	"model.layers.2.hc_ffn_base",
	"model.layers.2.hc_ffn_fn",
	"model.layers.2.hc_ffn_scale",
	"model.layers.2.input_layernorm",
	"model.layers.2.post_attention_layernorm",
	"model.layers.2.self_attn.A_log",
	"model.layers.2.self_attn.b_proj",
	"model.layers.2.self_attn.dt_bias",
	"model.layers.2.self_attn.f_a_proj",
	"model.layers.2.self_attn.f_b_proj",
	"model.layers.2.self_attn.fused_qkvbfg_a_proj",
	"model.layers.2.self_attn.g_a_proj",
	"model.layers.2.self_attn.g_b_proj",
	"model.layers.2.self_attn.k_conv1d",
	"model.layers.2.self_attn.k_proj",
	"model.layers.2.self_attn.o_norm",
	"model.layers.2.self_attn.o_proj",
	"model.layers.2.self_attn.q_conv1d",
	"model.layers.2.self_attn.q_proj",
	"model.layers.2.self_attn.qkv_proj",
	"model.layers.2.self_attn.v_conv1d",
	"model.layers.2.self_attn.v_proj",
	"model.layers.20.hc_attn_base",
	"model.layers.20.hc_attn_fn",
	"model.layers.20.hc_attn_scale",
	"model.layers.20.hc_ffn_base",
	"model.layers.20.hc_ffn_fn",
	"model.layers.20.hc_ffn_scale",
	"model.layers.20.input_layernorm",
	"model.layers.20.mlp.gate",
	"model.layers.20.mlp.gate.e_score_correction_bias",
	"model.layers.20.post_attention_layernorm",
	"model.layers.20.self_attn.A_log",
	"model.layers.20.self_attn.b_proj",
	"model.layers.20.self_attn.dt_bias",
	"model.layers.20.self_attn.f_a_proj",
	"model.layers.20.self_attn.f_b_proj",
	"model.layers.20.self_attn.fused_qkvbfg_a_proj",
	"model.layers.20.self_attn.g_a_proj",
	"model.layers.20.self_attn.g_b_proj",
	"model.layers.20.self_attn.k_conv1d",
	"model.layers.20.self_attn.k_proj",
	"model.layers.20.self_attn.o_norm",
	"model.layers.20.self_attn.o_proj",
	"model.layers.20.self_attn.q_conv1d",
	"model.layers.20.self_attn.q_proj",
	"model.layers.20.self_attn.qkv_proj",
	"model.layers.20.self_attn.v_conv1d",
	"model.layers.20.self_attn.v_proj",
	"model.layers.21.hc_attn_base",
	"model.layers.21.hc_attn_fn",
	"model.layers.21.hc_attn_scale",
	"model.layers.21.hc_ffn_base",
	"model.layers.21.hc_ffn_fn",
	"model.layers.21.hc_ffn_scale",
	"model.layers.21.input_layernorm",
	"model.layers.21.mlp.gate",
	"model.layers.21.mlp.gate.e_score_correction_bias",
	"model.layers.21.post_attention_layernorm",
	"model.layers.21.self_attn.A_log",
	"model.layers.21.self_attn.b_proj",
	"model.layers.21.self_attn.dt_bias",
	"model.layers.21.self_attn.f_a_proj",
	"model.layers.21.self_attn.f_b_proj",
	"model.layers.21.self_attn.fused_qkvbfg_a_proj",
	"model.layers.21.self_attn.g_a_proj",
	"model.layers.21.self_attn.g_b_proj",
	"model.layers.21.self_attn.k_conv1d",
	"model.layers.21.self_attn.k_proj",
	"model.layers.21.self_attn.o_norm",
	"model.layers.21.self_attn.o_proj",
	"model.layers.21.self_attn.q_conv1d",
	"model.layers.21.self_attn.q_proj",
	"model.layers.21.self_attn.qkv_proj",
	"model.layers.21.self_attn.v_conv1d",
	"model.layers.21.self_attn.v_proj",
	"model.layers.22.hc_attn_base",
	"model.layers.22.hc_attn_fn",
	"model.layers.22.hc_attn_scale",
	"model.layers.22.hc_ffn_base",
	"model.layers.22.hc_ffn_fn",
	"model.layers.22.hc_ffn_scale",
	"model.layers.22.input_layernorm",
	"model.layers.22.mlp.gate",
	"model.layers.22.mlp.gate.e_score_correction_bias",
	"model.layers.22.post_attention_layernorm",
	"model.layers.22.self_attn.A_log",
	"model.layers.22.self_attn.b_proj",
	"model.layers.22.self_attn.dt_bias",
	"model.layers.22.self_attn.f_a_proj",
	"model.layers.22.self_attn.f_b_proj",
	"model.layers.22.self_attn.fused_qkvbfg_a_proj",
	"model.layers.22.self_attn.g_a_proj",
	"model.layers.22.self_attn.g_b_proj",
	"model.layers.22.self_attn.k_conv1d",
	"model.layers.22.self_attn.k_proj",
	"model.layers.22.self_attn.o_norm",
	"model.layers.22.self_attn.o_proj",
	"model.layers.22.self_attn.q_conv1d",
	"model.layers.22.self_attn.q_proj",
	"model.layers.22.self_attn.qkv_proj",
	"model.layers.22.self_attn.v_conv1d",
	"model.layers.22.self_attn.v_proj",
	"model.layers.23.hc_attn_base",
	"model.layers.23.hc_attn_fn",
	"model.layers.23.hc_attn_scale",
	"model.layers.23.hc_ffn_base",
	"model.layers.23.hc_ffn_fn",
	"model.layers.23.hc_ffn_scale",
	"model.layers.23.input_layernorm",
	"model.layers.23.mlp.gate",
	"model.layers.23.mlp.gate.e_score_correction_bias",
	"model.layers.23.post_attention_layernorm",
	"model.layers.23.self_attn.indexer.index_kpool_compress_ape",
	"model.layers.23.self_attn.indexer.index_kpool_compress_gate",
	"model.layers.23.self_attn.indexer.k_norm",
	"model.layers.23.self_attn.indexer.k_norm.bias",
	"model.layers.23.self_attn.indexer.weights_proj",
	"model.layers.23.self_attn.indexer.wk",
	"model.layers.23.self_attn.indexer.wq_b",
	"model.layers.23.self_attn.kv_a_layernorm",
	"model.layers.23.self_attn.kv_b_proj",
	"model.layers.23.self_attn.q_a_layernorm",
	"model.layers.24.hc_attn_base",
	"model.layers.24.hc_attn_fn",
	"model.layers.24.hc_attn_scale",
	"model.layers.24.hc_ffn_base",
	"model.layers.24.hc_ffn_fn",
	"model.layers.24.hc_ffn_scale",
	"model.layers.24.input_layernorm",
	"model.layers.24.mlp.gate",
	"model.layers.24.mlp.gate.e_score_correction_bias",
	"model.layers.24.post_attention_layernorm",
	"model.layers.24.self_attn.A_log",
	"model.layers.24.self_attn.b_proj",
	"model.layers.24.self_attn.dt_bias",
	"model.layers.24.self_attn.f_a_proj",
	"model.layers.24.self_attn.f_b_proj",
	"model.layers.24.self_attn.fused_qkvbfg_a_proj",
	"model.layers.24.self_attn.g_a_proj",
	"model.layers.24.self_attn.g_b_proj",
	"model.layers.24.self_attn.k_conv1d",
	"model.layers.24.self_attn.k_proj",
	"model.layers.24.self_attn.o_norm",
	"model.layers.24.self_attn.o_proj",
	"model.layers.24.self_attn.q_conv1d",
	"model.layers.24.self_attn.q_proj",
	"model.layers.24.self_attn.qkv_proj",
	"model.layers.24.self_attn.v_conv1d",
	"model.layers.24.self_attn.v_proj",
	"model.layers.25.hc_attn_base",
	"model.layers.25.hc_attn_fn",
	"model.layers.25.hc_attn_scale",
	"model.layers.25.hc_ffn_base",
	"model.layers.25.hc_ffn_fn",
	"model.layers.25.hc_ffn_scale",
	"model.layers.25.input_layernorm",
	"model.layers.25.mlp.gate",
	"model.layers.25.mlp.gate.e_score_correction_bias",
	"model.layers.25.post_attention_layernorm",
	"model.layers.25.self_attn.A_log",
	"model.layers.25.self_attn.b_proj",
	"model.layers.25.self_attn.dt_bias",
	"model.layers.25.self_attn.f_a_proj",
	"model.layers.25.self_attn.f_b_proj",
	"model.layers.25.self_attn.fused_qkvbfg_a_proj",
	"model.layers.25.self_attn.g_a_proj",
	"model.layers.25.self_attn.g_b_proj",
	"model.layers.25.self_attn.k_conv1d",
	"model.layers.25.self_attn.k_proj",
	"model.layers.25.self_attn.o_norm",
	"model.layers.25.self_attn.o_proj",
	"model.layers.25.self_attn.q_conv1d",
	"model.layers.25.self_attn.q_proj",
	"model.layers.25.self_attn.qkv_proj",
	"model.layers.25.self_attn.v_conv1d",
	"model.layers.25.self_attn.v_proj",
	"model.layers.26.hc_attn_base",
	"model.layers.26.hc_attn_fn",
	"model.layers.26.hc_attn_scale",
	"model.layers.26.hc_ffn_base",
	"model.layers.26.hc_ffn_fn",
	"model.layers.26.hc_ffn_scale",
	"model.layers.26.input_layernorm",
	"model.layers.26.mlp.gate",
	"model.layers.26.mlp.gate.e_score_correction_bias",
	"model.layers.26.post_attention_layernorm",
	"model.layers.26.self_attn.A_log",
	"model.layers.26.self_attn.b_proj",
	"model.layers.26.self_attn.dt_bias",
	"model.layers.26.self_attn.f_a_proj",
	"model.layers.26.self_attn.f_b_proj",
	"model.layers.26.self_attn.fused_qkvbfg_a_proj",
	"model.layers.26.self_attn.g_a_proj",
	"model.layers.26.self_attn.g_b_proj",
	"model.layers.26.self_attn.k_conv1d",
	"model.layers.26.self_attn.k_proj",
	"model.layers.26.self_attn.o_norm",
	"model.layers.26.self_attn.o_proj",
	"model.layers.26.self_attn.q_conv1d",
	"model.layers.26.self_attn.q_proj",
	"model.layers.26.self_attn.qkv_proj",
	"model.layers.26.self_attn.v_conv1d",
	"model.layers.26.self_attn.v_proj",
	"model.layers.27.hc_attn_base",
	"model.layers.27.hc_attn_fn",
	"model.layers.27.hc_attn_scale",
	"model.layers.27.hc_ffn_base",
	"model.layers.27.hc_ffn_fn",
	"model.layers.27.hc_ffn_scale",
	"model.layers.27.input_layernorm",
	"model.layers.27.mlp.gate",
	"model.layers.27.mlp.gate.e_score_correction_bias",
	"model.layers.27.post_attention_layernorm",
	"model.layers.27.self_attn.indexer.index_kpool_compress_ape",
	"model.layers.27.self_attn.indexer.index_kpool_compress_gate",
	"model.layers.27.self_attn.indexer.k_norm",
	"model.layers.27.self_attn.indexer.k_norm.bias",
	"model.layers.27.self_attn.indexer.weights_proj",
	"model.layers.27.self_attn.indexer.wk",
	"model.layers.27.self_attn.indexer.wq_b",
	"model.layers.27.self_attn.kv_a_layernorm",
	"model.layers.27.self_attn.kv_b_proj",
	"model.layers.27.self_attn.q_a_layernorm",
	"model.layers.28.hc_attn_base",
	"model.layers.28.hc_attn_fn",
	"model.layers.28.hc_attn_scale",
	"model.layers.28.hc_ffn_base",
	"model.layers.28.hc_ffn_fn",
	"model.layers.28.hc_ffn_scale",
	"model.layers.28.input_layernorm",
	"model.layers.28.mlp.gate",
	"model.layers.28.mlp.gate.e_score_correction_bias",
	"model.layers.28.post_attention_layernorm",
	"model.layers.28.self_attn.A_log",
	"model.layers.28.self_attn.b_proj",
	"model.layers.28.self_attn.dt_bias",
	"model.layers.28.self_attn.f_a_proj",
	"model.layers.28.self_attn.f_b_proj",
	"model.layers.28.self_attn.fused_qkvbfg_a_proj",
	"model.layers.28.self_attn.g_a_proj",
	"model.layers.28.self_attn.g_b_proj",
	"model.layers.28.self_attn.k_conv1d",
	"model.layers.28.self_attn.k_proj",
	"model.layers.28.self_attn.o_norm",
	"model.layers.28.self_attn.o_proj",
	"model.layers.28.self_attn.q_conv1d",
	"model.layers.28.self_attn.q_proj",
	"model.layers.28.self_attn.qkv_proj",
	"model.layers.28.self_attn.v_conv1d",
	"model.layers.28.self_attn.v_proj",
	"model.layers.29.hc_attn_base",
	"model.layers.29.hc_attn_fn",
	"model.layers.29.hc_attn_scale",
	"model.layers.29.hc_ffn_base",
	"model.layers.29.hc_ffn_fn",
	"model.layers.29.hc_ffn_scale",
	"model.layers.29.input_layernorm",
	"model.layers.29.mlp.gate",
	"model.layers.29.mlp.gate.e_score_correction_bias",
	"model.layers.29.post_attention_layernorm",
	"model.layers.29.self_attn.A_log",
	"model.layers.29.self_attn.b_proj",
	"model.layers.29.self_attn.dt_bias",
	"model.layers.29.self_attn.f_a_proj",
	"model.layers.29.self_attn.f_b_proj",
	"model.layers.29.self_attn.fused_qkvbfg_a_proj",
	"model.layers.29.self_attn.g_a_proj",
	"model.layers.29.self_attn.g_b_proj",
	"model.layers.29.self_attn.k_conv1d",
	"model.layers.29.self_attn.k_proj",
	"model.layers.29.self_attn.o_norm",
	"model.layers.29.self_attn.o_proj",
	"model.layers.29.self_attn.q_conv1d",
	"model.layers.29.self_attn.q_proj",
	"model.layers.29.self_attn.qkv_proj",
	"model.layers.29.self_attn.v_conv1d",
	"model.layers.29.self_attn.v_proj",
	"model.layers.3.hc_attn_base",
	"model.layers.3.hc_attn_fn",
	"model.layers.3.hc_attn_scale",
	"model.layers.3.hc_ffn_base",
	"model.layers.3.hc_ffn_fn",
	"model.layers.3.hc_ffn_scale",
	"model.layers.3.input_layernorm",
	"model.layers.3.mlp.gate",
	"model.layers.3.mlp.gate.e_score_correction_bias",
	"model.layers.3.post_attention_layernorm",
	"model.layers.3.self_attn.indexer.index_kpool_compress_ape",
	"model.layers.3.self_attn.indexer.index_kpool_compress_gate",
	"model.layers.3.self_attn.indexer.k_norm",
	"model.layers.3.self_attn.indexer.k_norm.bias",
	"model.layers.3.self_attn.indexer.weights_proj",
	"model.layers.3.self_attn.indexer.wk",
	"model.layers.3.self_attn.indexer.wq_b",
	"model.layers.3.self_attn.kv_a_layernorm",
	"model.layers.3.self_attn.kv_b_proj",
	"model.layers.3.self_attn.q_a_layernorm",
	"model.layers.30.hc_attn_base",
	"model.layers.30.hc_attn_fn",
	"model.layers.30.hc_attn_scale",
	"model.layers.30.hc_ffn_base",
	"model.layers.30.hc_ffn_fn",
	"model.layers.30.hc_ffn_scale",
	"model.layers.30.input_layernorm",
	"model.layers.30.mlp.gate",
	"model.layers.30.mlp.gate.e_score_correction_bias",
	"model.layers.30.post_attention_layernorm",
	"model.layers.30.self_attn.A_log",
	"model.layers.30.self_attn.b_proj",
	"model.layers.30.self_attn.dt_bias",
	"model.layers.30.self_attn.f_a_proj",
	"model.layers.30.self_attn.f_b_proj",
	"model.layers.30.self_attn.fused_qkvbfg_a_proj",
	"model.layers.30.self_attn.g_a_proj",
	"model.layers.30.self_attn.g_b_proj",
	"model.layers.30.self_attn.k_conv1d",
	"model.layers.30.self_attn.k_proj",
	"model.layers.30.self_attn.o_norm",
	"model.layers.30.self_attn.o_proj",
	"model.layers.30.self_attn.q_conv1d",
	"model.layers.30.self_attn.q_proj",
	"model.layers.30.self_attn.qkv_proj",
	"model.layers.30.self_attn.v_conv1d",
	"model.layers.30.self_attn.v_proj",
	"model.layers.31.hc_attn_base",
	"model.layers.31.hc_attn_fn",
	"model.layers.31.hc_attn_scale",
	"model.layers.31.hc_ffn_base",
	"model.layers.31.hc_ffn_fn",
	"model.layers.31.hc_ffn_scale",
	"model.layers.31.input_layernorm",
	"model.layers.31.mlp.gate",
	"model.layers.31.mlp.gate.e_score_correction_bias",
	"model.layers.31.post_attention_layernorm",
	"model.layers.31.self_attn.indexer.index_kpool_compress_ape",
	"model.layers.31.self_attn.indexer.index_kpool_compress_gate",
	"model.layers.31.self_attn.indexer.k_norm",
	"model.layers.31.self_attn.indexer.k_norm.bias",
	"model.layers.31.self_attn.indexer.weights_proj",
	"model.layers.31.self_attn.indexer.wk",
	"model.layers.31.self_attn.indexer.wq_b",
	"model.layers.31.self_attn.kv_a_layernorm",
	"model.layers.31.self_attn.kv_b_proj",
	"model.layers.31.self_attn.q_a_layernorm",
	"model.layers.32.hc_attn_base",
	"model.layers.32.hc_attn_fn",
	"model.layers.32.hc_attn_scale",
	"model.layers.32.hc_ffn_base",
	"model.layers.32.hc_ffn_fn",
	"model.layers.32.hc_ffn_scale",
	"model.layers.32.input_layernorm",
	"model.layers.32.mlp.gate",
	"model.layers.32.mlp.gate.e_score_correction_bias",
	"model.layers.32.post_attention_layernorm",
	"model.layers.32.self_attn.A_log",
	"model.layers.32.self_attn.b_proj",
	"model.layers.32.self_attn.dt_bias",
	"model.layers.32.self_attn.f_a_proj",
	"model.layers.32.self_attn.f_b_proj",
	"model.layers.32.self_attn.fused_qkvbfg_a_proj",
	"model.layers.32.self_attn.g_a_proj",
	"model.layers.32.self_attn.g_b_proj",
	"model.layers.32.self_attn.k_conv1d",
	"model.layers.32.self_attn.k_proj",
	"model.layers.32.self_attn.o_norm",
	"model.layers.32.self_attn.o_proj",
	"model.layers.32.self_attn.q_conv1d",
	"model.layers.32.self_attn.q_proj",
	"model.layers.32.self_attn.qkv_proj",
	"model.layers.32.self_attn.v_conv1d",
	"model.layers.32.self_attn.v_proj",
	"model.layers.33.hc_attn_base",
	"model.layers.33.hc_attn_fn",
	"model.layers.33.hc_attn_scale",
	"model.layers.33.hc_ffn_base",
	"model.layers.33.hc_ffn_fn",
	"model.layers.33.hc_ffn_scale",
	"model.layers.33.input_layernorm",
	"model.layers.33.mlp.gate",
	"model.layers.33.mlp.gate.e_score_correction_bias",
	"model.layers.33.post_attention_layernorm",
	"model.layers.33.self_attn.A_log",
	"model.layers.33.self_attn.b_proj",
	"model.layers.33.self_attn.dt_bias",
	"model.layers.33.self_attn.f_a_proj",
	"model.layers.33.self_attn.f_b_proj",
	"model.layers.33.self_attn.fused_qkvbfg_a_proj",
	"model.layers.33.self_attn.g_a_proj",
	"model.layers.33.self_attn.g_b_proj",
	"model.layers.33.self_attn.k_conv1d",
	"model.layers.33.self_attn.k_proj",
	"model.layers.33.self_attn.o_norm",
	"model.layers.33.self_attn.o_proj",
	"model.layers.33.self_attn.q_conv1d",
	"model.layers.33.self_attn.q_proj",
	"model.layers.33.self_attn.qkv_proj",
	"model.layers.33.self_attn.v_conv1d",
	"model.layers.33.self_attn.v_proj",
	"model.layers.34.hc_attn_base",
	"model.layers.34.hc_attn_fn",
	"model.layers.34.hc_attn_scale",
	"model.layers.34.hc_ffn_base",
	"model.layers.34.hc_ffn_fn",
	"model.layers.34.hc_ffn_scale",
	"model.layers.34.input_layernorm",
	"model.layers.34.mlp.gate",
	"model.layers.34.mlp.gate.e_score_correction_bias",
	"model.layers.34.post_attention_layernorm",
	"model.layers.34.self_attn.A_log",
	"model.layers.34.self_attn.b_proj",
	"model.layers.34.self_attn.dt_bias",
	"model.layers.34.self_attn.f_a_proj",
	"model.layers.34.self_attn.f_b_proj",
	"model.layers.34.self_attn.fused_qkvbfg_a_proj",
	"model.layers.34.self_attn.g_a_proj",
	"model.layers.34.self_attn.g_b_proj",
	"model.layers.34.self_attn.k_conv1d",
	"model.layers.34.self_attn.k_proj",
	"model.layers.34.self_attn.o_norm",
	"model.layers.34.self_attn.o_proj",
	"model.layers.34.self_attn.q_conv1d",
	"model.layers.34.self_attn.q_proj",
	"model.layers.34.self_attn.qkv_proj",
	"model.layers.34.self_attn.v_conv1d",
	"model.layers.34.self_attn.v_proj",
	"model.layers.35.hc_attn_base",
	"model.layers.35.hc_attn_fn",
	"model.layers.35.hc_attn_scale",
	"model.layers.35.hc_ffn_base",
	"model.layers.35.hc_ffn_fn",
	"model.layers.35.hc_ffn_scale",
	"model.layers.35.input_layernorm",
	"model.layers.35.mlp.gate",
	"model.layers.35.mlp.gate.e_score_correction_bias",
	"model.layers.35.post_attention_layernorm",
	"model.layers.35.self_attn.indexer.index_kpool_compress_ape",
	"model.layers.35.self_attn.indexer.index_kpool_compress_gate",
	"model.layers.35.self_attn.indexer.k_norm",
	"model.layers.35.self_attn.indexer.k_norm.bias",
	"model.layers.35.self_attn.indexer.weights_proj",
	"model.layers.35.self_attn.indexer.wk",
	"model.layers.35.self_attn.indexer.wq_b",
	"model.layers.35.self_attn.kv_a_layernorm",
	"model.layers.35.self_attn.kv_b_proj",
	"model.layers.35.self_attn.q_a_layernorm",
	"model.layers.36.hc_attn_base",
	"model.layers.36.hc_attn_fn",
	"model.layers.36.hc_attn_scale",
	"model.layers.36.hc_ffn_base",
	"model.layers.36.hc_ffn_fn",
	"model.layers.36.hc_ffn_scale",
	"model.layers.36.input_layernorm",
	"model.layers.36.mlp.gate",
	"model.layers.36.mlp.gate.e_score_correction_bias",
	"model.layers.36.post_attention_layernorm",
	"model.layers.36.self_attn.A_log",
	"model.layers.36.self_attn.b_proj",
	"model.layers.36.self_attn.dt_bias",
	"model.layers.36.self_attn.f_a_proj",
	"model.layers.36.self_attn.f_b_proj",
	"model.layers.36.self_attn.fused_qkvbfg_a_proj",
	"model.layers.36.self_attn.g_a_proj",
	"model.layers.36.self_attn.g_b_proj",
	"model.layers.36.self_attn.k_conv1d",
	"model.layers.36.self_attn.k_proj",
	"model.layers.36.self_attn.o_norm",
	"model.layers.36.self_attn.o_proj",
	"model.layers.36.self_attn.q_conv1d",
	"model.layers.36.self_attn.q_proj",
	"model.layers.36.self_attn.qkv_proj",
	"model.layers.36.self_attn.v_conv1d",
	"model.layers.36.self_attn.v_proj",
	"model.layers.37.hc_attn_base",
	"model.layers.37.hc_attn_fn",
	"model.layers.37.hc_attn_scale",
	"model.layers.37.hc_ffn_base",
	"model.layers.37.hc_ffn_fn",
	"model.layers.37.hc_ffn_scale",
	"model.layers.37.input_layernorm",
	"model.layers.37.mlp.gate",
	"model.layers.37.mlp.gate.e_score_correction_bias",
	"model.layers.37.post_attention_layernorm",
	"model.layers.37.self_attn.A_log",
	"model.layers.37.self_attn.b_proj",
	"model.layers.37.self_attn.dt_bias",
	"model.layers.37.self_attn.f_a_proj",
	"model.layers.37.self_attn.f_b_proj",
	"model.layers.37.self_attn.fused_qkvbfg_a_proj",
	"model.layers.37.self_attn.g_a_proj",
	"model.layers.37.self_attn.g_b_proj",
	"model.layers.37.self_attn.k_conv1d",
	"model.layers.37.self_attn.k_proj",
	"model.layers.37.self_attn.o_norm",
	"model.layers.37.self_attn.o_proj",
	"model.layers.37.self_attn.q_conv1d",
	"model.layers.37.self_attn.q_proj",
	"model.layers.37.self_attn.qkv_proj",
	"model.layers.37.self_attn.v_conv1d",
	"model.layers.37.self_attn.v_proj",
	"model.layers.38.hc_attn_base",
	"model.layers.38.hc_attn_fn",
	"model.layers.38.hc_attn_scale",
	"model.layers.38.hc_ffn_base",
	"model.layers.38.hc_ffn_fn",
	"model.layers.38.hc_ffn_scale",
	"model.layers.38.input_layernorm",
	"model.layers.38.mlp.gate",
	"model.layers.38.mlp.gate.e_score_correction_bias",
	"model.layers.38.post_attention_layernorm",
	"model.layers.38.self_attn.A_log",
	"model.layers.38.self_attn.b_proj",
	"model.layers.38.self_attn.dt_bias",
	"model.layers.38.self_attn.f_a_proj",
	"model.layers.38.self_attn.f_b_proj",
	"model.layers.38.self_attn.fused_qkvbfg_a_proj",
	"model.layers.38.self_attn.g_a_proj",
	"model.layers.38.self_attn.g_b_proj",
	"model.layers.38.self_attn.k_conv1d",
	"model.layers.38.self_attn.k_proj",
	"model.layers.38.self_attn.o_norm",
	"model.layers.38.self_attn.o_proj",
	"model.layers.38.self_attn.q_conv1d",
	"model.layers.38.self_attn.q_proj",
	"model.layers.38.self_attn.qkv_proj",
	"model.layers.38.self_attn.v_conv1d",
	"model.layers.38.self_attn.v_proj",
	"model.layers.39.hc_attn_base",
	"model.layers.39.hc_attn_fn",
	"model.layers.39.hc_attn_scale",
	"model.layers.39.hc_ffn_base",
	"model.layers.39.hc_ffn_fn",
	"model.layers.39.hc_ffn_scale",
	"model.layers.39.input_layernorm",
	"model.layers.39.mlp.gate",
	"model.layers.39.mlp.gate.e_score_correction_bias",
	"model.layers.39.post_attention_layernorm",
	"model.layers.39.self_attn.indexer.index_kpool_compress_ape",
	"model.layers.39.self_attn.indexer.index_kpool_compress_gate",
	"model.layers.39.self_attn.indexer.k_norm",
	"model.layers.39.self_attn.indexer.k_norm.bias",
	"model.layers.39.self_attn.indexer.weights_proj",
	"model.layers.39.self_attn.indexer.wk",
	"model.layers.39.self_attn.indexer.wq_b",
	"model.layers.39.self_attn.kv_a_layernorm",
	"model.layers.39.self_attn.kv_b_proj",
	"model.layers.39.self_attn.q_a_layernorm",
	"model.layers.4.hc_attn_base",
	"model.layers.4.hc_attn_fn",
	"model.layers.4.hc_attn_scale",
	"model.layers.4.hc_ffn_base",
	"model.layers.4.hc_ffn_fn",
	"model.layers.4.hc_ffn_scale",
	"model.layers.4.input_layernorm",
	"model.layers.4.mlp.gate",
	"model.layers.4.mlp.gate.e_score_correction_bias",
	"model.layers.4.post_attention_layernorm",
	"model.layers.4.self_attn.A_log",
	"model.layers.4.self_attn.b_proj",
	"model.layers.4.self_attn.dt_bias",
	"model.layers.4.self_attn.f_a_proj",
	"model.layers.4.self_attn.f_b_proj",
	"model.layers.4.self_attn.fused_qkvbfg_a_proj",
	"model.layers.4.self_attn.g_a_proj",
	"model.layers.4.self_attn.g_b_proj",
	"model.layers.4.self_attn.k_conv1d",
	"model.layers.4.self_attn.k_proj",
	"model.layers.4.self_attn.o_norm",
	"model.layers.4.self_attn.o_proj",
	"model.layers.4.self_attn.q_conv1d",
	"model.layers.4.self_attn.q_proj",
	"model.layers.4.self_attn.qkv_proj",
	"model.layers.4.self_attn.v_conv1d",
	"model.layers.4.self_attn.v_proj",
	"model.layers.40.hc_attn_base",
	"model.layers.40.hc_attn_fn",
	"model.layers.40.hc_attn_scale",
	"model.layers.40.hc_ffn_base",
	"model.layers.40.hc_ffn_fn",
	"model.layers.40.hc_ffn_scale",
	"model.layers.40.input_layernorm",
	"model.layers.40.mlp.gate",
	"model.layers.40.mlp.gate.e_score_correction_bias",
	"model.layers.40.post_attention_layernorm",
	"model.layers.40.self_attn.A_log",
	"model.layers.40.self_attn.b_proj",
	"model.layers.40.self_attn.dt_bias",
	"model.layers.40.self_attn.f_a_proj",
	"model.layers.40.self_attn.f_b_proj",
	"model.layers.40.self_attn.fused_qkvbfg_a_proj",
	"model.layers.40.self_attn.g_a_proj",
	"model.layers.40.self_attn.g_b_proj",
	"model.layers.40.self_attn.k_conv1d",
	"model.layers.40.self_attn.k_proj",
	"model.layers.40.self_attn.o_norm",
	"model.layers.40.self_attn.o_proj",
	"model.layers.40.self_attn.q_conv1d",
	"model.layers.40.self_attn.q_proj",
	"model.layers.40.self_attn.qkv_proj",
	"model.layers.40.self_attn.v_conv1d",
	"model.layers.40.self_attn.v_proj",
	"model.layers.41.hc_attn_base",
	"model.layers.41.hc_attn_fn",
	"model.layers.41.hc_attn_scale",
	"model.layers.41.hc_ffn_base",
	"model.layers.41.hc_ffn_fn",
	"model.layers.41.hc_ffn_scale",
	"model.layers.41.input_layernorm",
	"model.layers.41.mlp.gate",
	"model.layers.41.mlp.gate.e_score_correction_bias",
	"model.layers.41.post_attention_layernorm",
	"model.layers.41.self_attn.A_log",
	"model.layers.41.self_attn.b_proj",
	"model.layers.41.self_attn.dt_bias",
	"model.layers.41.self_attn.f_a_proj",
	"model.layers.41.self_attn.f_b_proj",
	"model.layers.41.self_attn.fused_qkvbfg_a_proj",
	"model.layers.41.self_attn.g_a_proj",
	"model.layers.41.self_attn.g_b_proj",
	"model.layers.41.self_attn.k_conv1d",
	"model.layers.41.self_attn.k_proj",
	"model.layers.41.self_attn.o_norm",
	"model.layers.41.self_attn.o_proj",
	"model.layers.41.self_attn.q_conv1d",
	"model.layers.41.self_attn.q_proj",
	"model.layers.41.self_attn.qkv_proj",
	"model.layers.41.self_attn.v_conv1d",
	"model.layers.41.self_attn.v_proj",
	"model.layers.42.hc_attn_base",
	"model.layers.42.hc_attn_fn",
	"model.layers.42.hc_attn_scale",
	"model.layers.42.hc_ffn_base",
	"model.layers.42.hc_ffn_fn",
	"model.layers.42.hc_ffn_scale",
	"model.layers.42.input_layernorm",
	"model.layers.42.mlp.gate",
	"model.layers.42.mlp.gate.e_score_correction_bias",
	"model.layers.42.post_attention_layernorm",
	"model.layers.42.self_attn.A_log",
	"model.layers.42.self_attn.b_proj",
	"model.layers.42.self_attn.dt_bias",
	"model.layers.42.self_attn.f_a_proj",
	"model.layers.42.self_attn.f_b_proj",
	"model.layers.42.self_attn.fused_qkvbfg_a_proj",
	"model.layers.42.self_attn.g_a_proj",
	"model.layers.42.self_attn.g_b_proj",
	"model.layers.42.self_attn.k_conv1d",
	"model.layers.42.self_attn.k_proj",
	"model.layers.42.self_attn.o_norm",
	"model.layers.42.self_attn.o_proj",
	"model.layers.42.self_attn.q_conv1d",
	"model.layers.42.self_attn.q_proj",
	"model.layers.42.self_attn.qkv_proj",
	"model.layers.42.self_attn.v_conv1d",
	"model.layers.42.self_attn.v_proj",
	"model.layers.43.hc_attn_base",
	"model.layers.43.hc_attn_fn",
	"model.layers.43.hc_attn_scale",
	"model.layers.43.hc_ffn_base",
	"model.layers.43.hc_ffn_fn",
	"model.layers.43.hc_ffn_scale",
	"model.layers.43.input_layernorm",
	"model.layers.43.mlp.gate",
	"model.layers.43.mlp.gate.e_score_correction_bias",
	"model.layers.43.post_attention_layernorm",
	"model.layers.43.self_attn.indexer.index_kpool_compress_ape",
	"model.layers.43.self_attn.indexer.index_kpool_compress_gate",
	"model.layers.43.self_attn.indexer.k_norm",
	"model.layers.43.self_attn.indexer.k_norm.bias",
	"model.layers.43.self_attn.indexer.weights_proj",
	"model.layers.43.self_attn.indexer.wk",
	"model.layers.43.self_attn.indexer.wq_b",
	"model.layers.43.self_attn.kv_a_layernorm",
	"model.layers.43.self_attn.kv_b_proj",
	"model.layers.43.self_attn.q_a_layernorm",
	"model.layers.44.hc_attn_base",
	"model.layers.44.hc_attn_fn",
	"model.layers.44.hc_attn_scale",
	"model.layers.44.hc_ffn_base",
	"model.layers.44.hc_ffn_fn",
	"model.layers.44.hc_ffn_scale",
	"model.layers.44.input_layernorm",
	"model.layers.44.mlp.gate",
	"model.layers.44.mlp.gate.e_score_correction_bias",
	"model.layers.44.post_attention_layernorm",
	"model.layers.44.self_attn.A_log",
	"model.layers.44.self_attn.b_proj",
	"model.layers.44.self_attn.dt_bias",
	"model.layers.44.self_attn.f_a_proj",
	"model.layers.44.self_attn.f_b_proj",
	"model.layers.44.self_attn.fused_qkvbfg_a_proj",
	"model.layers.44.self_attn.g_a_proj",
	"model.layers.44.self_attn.g_b_proj",
	"model.layers.44.self_attn.k_conv1d",
	"model.layers.44.self_attn.k_proj",
	"model.layers.44.self_attn.o_norm",
	"model.layers.44.self_attn.o_proj",
	"model.layers.44.self_attn.q_conv1d",
	"model.layers.44.self_attn.q_proj",
	"model.layers.44.self_attn.qkv_proj",
	"model.layers.44.self_attn.v_conv1d",
	"model.layers.44.self_attn.v_proj",
	"model.layers.45.eh_proj",
	"model.layers.45.enorm",
	"model.layers.45.hnorm",
	"model.layers.45.input_layernorm",
	"model.layers.45.mlp.gate",
	"model.layers.45.mlp.gate.e_score_correction_bias",
	"model.layers.45.post_attention_layernorm",
	"model.layers.45.self_attn.indexer.index_kpool_compress_ape",
	"model.layers.45.self_attn.indexer.index_kpool_compress_gate",
	"model.layers.45.self_attn.indexer.k_norm",
	"model.layers.45.self_attn.indexer.k_norm.bias",
	"model.layers.45.self_attn.indexer.weights_proj",
	"model.layers.45.self_attn.indexer.wk",
	"model.layers.45.self_attn.indexer.wq_b",
	"model.layers.45.self_attn.kv_a_layernorm",
	"model.layers.45.self_attn.kv_b_proj",
	"model.layers.45.self_attn.q_a_layernorm",
	"model.layers.45.shared_head.norm",
	"model.layers.5.hc_attn_base",
	"model.layers.5.hc_attn_fn",
	"model.layers.5.hc_attn_scale",
	"model.layers.5.hc_ffn_base",
	"model.layers.5.hc_ffn_fn",
	"model.layers.5.hc_ffn_scale",
	"model.layers.5.input_layernorm",
	"model.layers.5.mlp.gate",
	"model.layers.5.mlp.gate.e_score_correction_bias",
	"model.layers.5.post_attention_layernorm",
	"model.layers.5.self_attn.A_log",
	"model.layers.5.self_attn.b_proj",
	"model.layers.5.self_attn.dt_bias",
	"model.layers.5.self_attn.f_a_proj",
	"model.layers.5.self_attn.f_b_proj",
	"model.layers.5.self_attn.fused_qkvbfg_a_proj",
	"model.layers.5.self_attn.g_a_proj",
	"model.layers.5.self_attn.g_b_proj",
	"model.layers.5.self_attn.k_conv1d",
	"model.layers.5.self_attn.k_proj",
	"model.layers.5.self_attn.o_norm",
	"model.layers.5.self_attn.o_proj",
	"model.layers.5.self_attn.q_conv1d",
	"model.layers.5.self_attn.q_proj",
	"model.layers.5.self_attn.qkv_proj",
	"model.layers.5.self_attn.v_conv1d",
	"model.layers.5.self_attn.v_proj",
	"model.layers.6.hc_attn_base",
	"model.layers.6.hc_attn_fn",
	"model.layers.6.hc_attn_scale",
	"model.layers.6.hc_ffn_base",
	"model.layers.6.hc_ffn_fn",
	"model.layers.6.hc_ffn_scale",
	"model.layers.6.input_layernorm",
	"model.layers.6.mlp.gate",
	"model.layers.6.mlp.gate.e_score_correction_bias",
	"model.layers.6.post_attention_layernorm",
	"model.layers.6.self_attn.A_log",
	"model.layers.6.self_attn.b_proj",
	"model.layers.6.self_attn.dt_bias",
	"model.layers.6.self_attn.f_a_proj",
	"model.layers.6.self_attn.f_b_proj",
	"model.layers.6.self_attn.fused_qkvbfg_a_proj",
	"model.layers.6.self_attn.g_a_proj",
	"model.layers.6.self_attn.g_b_proj",
	"model.layers.6.self_attn.k_conv1d",
	"model.layers.6.self_attn.k_proj",
	"model.layers.6.self_attn.o_norm",
	"model.layers.6.self_attn.o_proj",
	"model.layers.6.self_attn.q_conv1d",
	"model.layers.6.self_attn.q_proj",
	"model.layers.6.self_attn.qkv_proj",
	"model.layers.6.self_attn.v_conv1d",
	"model.layers.6.self_attn.v_proj",
	"model.layers.7.hc_attn_base",
	"model.layers.7.hc_attn_fn",
	"model.layers.7.hc_attn_scale",
	"model.layers.7.hc_ffn_base",
	"model.layers.7.hc_ffn_fn",
	"model.layers.7.hc_ffn_scale",
	"model.layers.7.input_layernorm",
	"model.layers.7.mlp.gate",
	"model.layers.7.mlp.gate.e_score_correction_bias",
	"model.layers.7.post_attention_layernorm",
	"model.layers.7.self_attn.indexer.index_kpool_compress_ape",
	"model.layers.7.self_attn.indexer.index_kpool_compress_gate",
	"model.layers.7.self_attn.indexer.k_norm",
	"model.layers.7.self_attn.indexer.k_norm.bias",
	"model.layers.7.self_attn.indexer.weights_proj",
	"model.layers.7.self_attn.indexer.wk",
	"model.layers.7.self_attn.indexer.wq_b",
	"model.layers.7.self_attn.kv_a_layernorm",
	"model.layers.7.self_attn.kv_b_proj",
	"model.layers.7.self_attn.q_a_layernorm",
	"model.layers.8.hc_attn_base",
	"model.layers.8.hc_attn_fn",
	"model.layers.8.hc_attn_scale",
	"model.layers.8.hc_ffn_base",
	"model.layers.8.hc_ffn_fn",
	"model.layers.8.hc_ffn_scale",
	"model.layers.8.input_layernorm",
	"model.layers.8.mlp.gate",
	"model.layers.8.mlp.gate.e_score_correction_bias",
	"model.layers.8.post_attention_layernorm",
	"model.layers.8.self_attn.A_log",
	"model.layers.8.self_attn.b_proj",
	"model.layers.8.self_attn.dt_bias",
	"model.layers.8.self_attn.f_a_proj",
	"model.layers.8.self_attn.f_b_proj",
	"model.layers.8.self_attn.fused_qkvbfg_a_proj",
	"model.layers.8.self_attn.g_a_proj",
	"model.layers.8.self_attn.g_b_proj",
	"model.layers.8.self_attn.k_conv1d",
	"model.layers.8.self_attn.k_proj",
	"model.layers.8.self_attn.o_norm",
	"model.layers.8.self_attn.o_proj",
	"model.layers.8.self_attn.q_conv1d",
	"model.layers.8.self_attn.q_proj",
	"model.layers.8.self_attn.qkv_proj",
	"model.layers.8.self_attn.v_conv1d",
	"model.layers.8.self_attn.v_proj",
	"model.layers.9.hc_attn_base",
	"model.layers.9.hc_attn_fn",
	"model.layers.9.hc_attn_scale",
	"model.layers.9.hc_ffn_base",
	"model.layers.9.hc_ffn_fn",
	"model.layers.9.hc_ffn_scale",
	"model.layers.9.input_layernorm",
	"model.layers.9.mlp.gate",
	"model.layers.9.mlp.gate.e_score_correction_bias",
	"model.layers.9.post_attention_layernorm",
	"model.layers.9.self_attn.A_log",
	"model.layers.9.self_attn.b_proj",
	"model.layers.9.self_attn.dt_bias",
	"model.layers.9.self_attn.f_a_proj",
	"model.layers.9.self_attn.f_b_proj",
	"model.layers.9.self_attn.fused_qkvbfg_a_proj",
	"model.layers.9.self_attn.g_a_proj",
	"model.layers.9.self_attn.g_b_proj",
	"model.layers.9.self_attn.k_conv1d",
	"model.layers.9.self_attn.k_proj",
	"model.layers.9.self_attn.o_norm",
	"model.layers.9.self_attn.o_proj",
	"model.layers.9.self_attn.q_conv1d",
	"model.layers.9.self_attn.q_proj",
	"model.layers.9.self_attn.qkv_proj",
	"model.layers.9.self_attn.v_conv1d",
	"model.layers.9.self_attn.v_proj",
	"model.norm",
	"model.visual",
	"router",
	"visual",
	"visual.blocks.0.attn.k_norm",
	"visual.blocks.0.attn.proj",
	"visual.blocks.0.attn.proj.bias",
	"visual.blocks.0.attn.q_norm",
	"visual.blocks.0.attn.qkv",
	"visual.blocks.0.attn.qkv.bias",
	"visual.blocks.0.mlp.down_proj",
	"visual.blocks.0.mlp.down_proj.bias",
	"visual.blocks.0.mlp.gate_proj",
	"visual.blocks.0.mlp.gate_proj.bias",
	"visual.blocks.0.mlp.up_proj",
	"visual.blocks.0.mlp.up_proj.bias",
	"visual.blocks.0.norm1",
	"visual.blocks.0.norm2",
	"visual.blocks.1.attn.k_norm",
	"visual.blocks.1.attn.proj",
	"visual.blocks.1.attn.proj.bias",
	"visual.blocks.1.attn.q_norm",
	"visual.blocks.1.attn.qkv",
	"visual.blocks.1.attn.qkv.bias",
	"visual.blocks.1.mlp.down_proj",
	"visual.blocks.1.mlp.down_proj.bias",
	"visual.blocks.1.mlp.gate_proj",
	"visual.blocks.1.mlp.gate_proj.bias",
	"visual.blocks.1.mlp.up_proj",
	"visual.blocks.1.mlp.up_proj.bias",
	"visual.blocks.1.norm1",
	"visual.blocks.1.norm2",
	"visual.blocks.10.attn.k_norm",
	"visual.blocks.10.attn.proj",
	"visual.blocks.10.attn.proj.bias",
	"visual.blocks.10.attn.q_norm",
	"visual.blocks.10.attn.qkv",
	"visual.blocks.10.attn.qkv.bias",
	"visual.blocks.10.mlp.down_proj",
	"visual.blocks.10.mlp.down_proj.bias",
	"visual.blocks.10.mlp.gate_proj",
	"visual.blocks.10.mlp.gate_proj.bias",
	"visual.blocks.10.mlp.up_proj",
	"visual.blocks.10.mlp.up_proj.bias",
	"visual.blocks.10.norm1",
	"visual.blocks.10.norm2",
	"visual.blocks.11.attn.k_norm",
	"visual.blocks.11.attn.proj",
	"visual.blocks.11.attn.proj.bias",
	"visual.blocks.11.attn.q_norm",
	"visual.blocks.11.attn.qkv",
	"visual.blocks.11.attn.qkv.bias",
	"visual.blocks.11.mlp.down_proj",
	"visual.blocks.11.mlp.down_proj.bias",
	"visual.blocks.11.mlp.gate_proj",
	"visual.blocks.11.mlp.gate_proj.bias",
	"visual.blocks.11.mlp.up_proj",
	"visual.blocks.11.mlp.up_proj.bias",
	"visual.blocks.11.norm1",
	"visual.blocks.11.norm2",
	"visual.blocks.12.attn.k_norm",
	"visual.blocks.12.attn.proj",
	"visual.blocks.12.attn.proj.bias",
	"visual.blocks.12.attn.q_norm",
	"visual.blocks.12.attn.qkv",
	"visual.blocks.12.attn.qkv.bias",
	"visual.blocks.12.mlp.down_proj",
	"visual.blocks.12.mlp.down_proj.bias",
	"visual.blocks.12.mlp.gate_proj",
	"visual.blocks.12.mlp.gate_proj.bias",
	"visual.blocks.12.mlp.up_proj",
	"visual.blocks.12.mlp.up_proj.bias",
	"visual.blocks.12.norm1",
	"visual.blocks.12.norm2",
	"visual.blocks.13.attn.k_norm",
	"visual.blocks.13.attn.proj",
	"visual.blocks.13.attn.proj.bias",
	"visual.blocks.13.attn.q_norm",
	"visual.blocks.13.attn.qkv",
	"visual.blocks.13.attn.qkv.bias",
	"visual.blocks.13.mlp.down_proj",
	"visual.blocks.13.mlp.down_proj.bias",
	"visual.blocks.13.mlp.gate_proj",
	"visual.blocks.13.mlp.gate_proj.bias",
	"visual.blocks.13.mlp.up_proj",
	"visual.blocks.13.mlp.up_proj.bias",
	"visual.blocks.13.norm1",
	"visual.blocks.13.norm2",
	"visual.blocks.14.attn.k_norm",
	"visual.blocks.14.attn.proj",
	"visual.blocks.14.attn.proj.bias",
	"visual.blocks.14.attn.q_norm",
	"visual.blocks.14.attn.qkv",
	"visual.blocks.14.attn.qkv.bias",
	"visual.blocks.14.mlp.down_proj",
	"visual.blocks.14.mlp.down_proj.bias",
	"visual.blocks.14.mlp.gate_proj",
	"visual.blocks.14.mlp.gate_proj.bias",
	"visual.blocks.14.mlp.up_proj",
	"visual.blocks.14.mlp.up_proj.bias",
	"visual.blocks.14.norm1",
	"visual.blocks.14.norm2",
	"visual.blocks.15.attn.k_norm",
	"visual.blocks.15.attn.proj",
	"visual.blocks.15.attn.proj.bias",
	"visual.blocks.15.attn.q_norm",
	"visual.blocks.15.attn.qkv",
	"visual.blocks.15.attn.qkv.bias",
	"visual.blocks.15.mlp.down_proj",
	"visual.blocks.15.mlp.down_proj.bias",
	"visual.blocks.15.mlp.gate_proj",
	"visual.blocks.15.mlp.gate_proj.bias",
	"visual.blocks.15.mlp.up_proj",
	"visual.blocks.15.mlp.up_proj.bias",
	"visual.blocks.15.norm1",
	"visual.blocks.15.norm2",
	"visual.blocks.16.attn.k_norm",
	"visual.blocks.16.attn.proj",
	"visual.blocks.16.attn.proj.bias",
	"visual.blocks.16.attn.q_norm",
	"visual.blocks.16.attn.qkv",
	"visual.blocks.16.attn.qkv.bias",
	"visual.blocks.16.mlp.down_proj",
	"visual.blocks.16.mlp.down_proj.bias",
	"visual.blocks.16.mlp.gate_proj",
	"visual.blocks.16.mlp.gate_proj.bias",
	"visual.blocks.16.mlp.up_proj",
	"visual.blocks.16.mlp.up_proj.bias",
	"visual.blocks.16.norm1",
	"visual.blocks.16.norm2",
	"visual.blocks.17.attn.k_norm",
	"visual.blocks.17.attn.proj",
	"visual.blocks.17.attn.proj.bias",
	"visual.blocks.17.attn.q_norm",
	"visual.blocks.17.attn.qkv",
	"visual.blocks.17.attn.qkv.bias",
	"visual.blocks.17.mlp.down_proj",
	"visual.blocks.17.mlp.down_proj.bias",
	"visual.blocks.17.mlp.gate_proj",
	"visual.blocks.17.mlp.gate_proj.bias",
	"visual.blocks.17.mlp.up_proj",
	"visual.blocks.17.mlp.up_proj.bias",
	"visual.blocks.17.norm1",
	"visual.blocks.17.norm2",
	"visual.blocks.18.attn.k_norm",
	"visual.blocks.18.attn.proj",
	"visual.blocks.18.attn.proj.bias",
	"visual.blocks.18.attn.q_norm",
	"visual.blocks.18.attn.qkv",
	"visual.blocks.18.attn.qkv.bias",
	"visual.blocks.18.mlp.down_proj",
	"visual.blocks.18.mlp.down_proj.bias",
	"visual.blocks.18.mlp.gate_proj",
	"visual.blocks.18.mlp.gate_proj.bias",
	"visual.blocks.18.mlp.up_proj",
	"visual.blocks.18.mlp.up_proj.bias",
	"visual.blocks.18.norm1",
	"visual.blocks.18.norm2",
	"visual.blocks.19.attn.k_norm",
	"visual.blocks.19.attn.proj",
	"visual.blocks.19.attn.proj.bias",
	"visual.blocks.19.attn.q_norm",
	"visual.blocks.19.attn.qkv",
	"visual.blocks.19.attn.qkv.bias",
	"visual.blocks.19.mlp.down_proj",
	"visual.blocks.19.mlp.down_proj.bias",
	"visual.blocks.19.mlp.gate_proj",
	"visual.blocks.19.mlp.gate_proj.bias",
	"visual.blocks.19.mlp.up_proj",
	"visual.blocks.19.mlp.up_proj.bias",
	"visual.blocks.19.norm1",
	"visual.blocks.19.norm2",
	"visual.blocks.2.attn.k_norm",
	"visual.blocks.2.attn.proj",
	"visual.blocks.2.attn.proj.bias",
	"visual.blocks.2.attn.q_norm",
	"visual.blocks.2.attn.qkv",
	"visual.blocks.2.attn.qkv.bias",
	"visual.blocks.2.mlp.down_proj",
	"visual.blocks.2.mlp.down_proj.bias",
	"visual.blocks.2.mlp.gate_proj",
	"visual.blocks.2.mlp.gate_proj.bias",
	"visual.blocks.2.mlp.up_proj",
	"visual.blocks.2.mlp.up_proj.bias",
	"visual.blocks.2.norm1",
	"visual.blocks.2.norm2",
	"visual.blocks.20.attn.k_norm",
	"visual.blocks.20.attn.proj",
	"visual.blocks.20.attn.proj.bias",
	"visual.blocks.20.attn.q_norm",
	"visual.blocks.20.attn.qkv",
	"visual.blocks.20.attn.qkv.bias",
	"visual.blocks.20.mlp.down_proj",
	"visual.blocks.20.mlp.down_proj.bias",
	"visual.blocks.20.mlp.gate_proj",
	"visual.blocks.20.mlp.gate_proj.bias",
	"visual.blocks.20.mlp.up_proj",
	"visual.blocks.20.mlp.up_proj.bias",
	"visual.blocks.20.norm1",
	"visual.blocks.20.norm2",
	"visual.blocks.21.attn.k_norm",
	"visual.blocks.21.attn.proj",
	"visual.blocks.21.attn.proj.bias",
	"visual.blocks.21.attn.q_norm",
	"visual.blocks.21.attn.qkv",
	"visual.blocks.21.attn.qkv.bias",
	"visual.blocks.21.mlp.down_proj",
	"visual.blocks.21.mlp.down_proj.bias",
	"visual.blocks.21.mlp.gate_proj",
	"visual.blocks.21.mlp.gate_proj.bias",
	"visual.blocks.21.mlp.up_proj",
	"visual.blocks.21.mlp.up_proj.bias",
	"visual.blocks.21.norm1",
	"visual.blocks.21.norm2",
	"visual.blocks.22.attn.k_norm",
	"visual.blocks.22.attn.proj",
	"visual.blocks.22.attn.proj.bias",
	"visual.blocks.22.attn.q_norm",
	"visual.blocks.22.attn.qkv",
	"visual.blocks.22.attn.qkv.bias",
	"visual.blocks.22.mlp.down_proj",
	"visual.blocks.22.mlp.down_proj.bias",
	"visual.blocks.22.mlp.gate_proj",
	"visual.blocks.22.mlp.gate_proj.bias",
	"visual.blocks.22.mlp.up_proj",
	"visual.blocks.22.mlp.up_proj.bias",
	"visual.blocks.22.norm1",
	"visual.blocks.22.norm2",
	"visual.blocks.23.attn.k_norm",
	"visual.blocks.23.attn.proj",
	"visual.blocks.23.attn.proj.bias",
	"visual.blocks.23.attn.q_norm",
	"visual.blocks.23.attn.qkv",
	"visual.blocks.23.attn.qkv.bias",
	"visual.blocks.23.mlp.down_proj",
	"visual.blocks.23.mlp.down_proj.bias",
	"visual.blocks.23.mlp.gate_proj",
	"visual.blocks.23.mlp.gate_proj.bias",
	"visual.blocks.23.mlp.up_proj",
	"visual.blocks.23.mlp.up_proj.bias",
	"visual.blocks.23.norm1",
	"visual.blocks.23.norm2",
	"visual.blocks.3.attn.k_norm",
	"visual.blocks.3.attn.proj",
	"visual.blocks.3.attn.proj.bias",
	"visual.blocks.3.attn.q_norm",
	"visual.blocks.3.attn.qkv",
	"visual.blocks.3.attn.qkv.bias",
	"visual.blocks.3.mlp.down_proj",
	"visual.blocks.3.mlp.down_proj.bias",
	"visual.blocks.3.mlp.gate_proj",
	"visual.blocks.3.mlp.gate_proj.bias",
	"visual.blocks.3.mlp.up_proj",
	"visual.blocks.3.mlp.up_proj.bias",
	"visual.blocks.3.norm1",
	"visual.blocks.3.norm2",
	"visual.blocks.4.attn.k_norm",
	"visual.blocks.4.attn.proj",
	"visual.blocks.4.attn.proj.bias",
	"visual.blocks.4.attn.q_norm",
	"visual.blocks.4.attn.qkv",
	"visual.blocks.4.attn.qkv.bias",
	"visual.blocks.4.mlp.down_proj",
	"visual.blocks.4.mlp.down_proj.bias",
	"visual.blocks.4.mlp.gate_proj",
	"visual.blocks.4.mlp.gate_proj.bias",
	"visual.blocks.4.mlp.up_proj",
	"visual.blocks.4.mlp.up_proj.bias",
	"visual.blocks.4.norm1",
	"visual.blocks.4.norm2",
	"visual.blocks.5.attn.k_norm",
	"visual.blocks.5.attn.proj",
	"visual.blocks.5.attn.proj.bias",
	"visual.blocks.5.attn.q_norm",
	"visual.blocks.5.attn.qkv",
	"visual.blocks.5.attn.qkv.bias",
	"visual.blocks.5.mlp.down_proj",
	"visual.blocks.5.mlp.down_proj.bias",
	"visual.blocks.5.mlp.gate_proj",
	"visual.blocks.5.mlp.gate_proj.bias",
	"visual.blocks.5.mlp.up_proj",
	"visual.blocks.5.mlp.up_proj.bias",
	"visual.blocks.5.norm1",
	"visual.blocks.5.norm2",
	"visual.blocks.6.attn.k_norm",
	"visual.blocks.6.attn.proj",
	"visual.blocks.6.attn.proj.bias",
	"visual.blocks.6.attn.q_norm",
	"visual.blocks.6.attn.qkv",
	"visual.blocks.6.attn.qkv.bias",
	"visual.blocks.6.mlp.down_proj",
	"visual.blocks.6.mlp.down_proj.bias",
	"visual.blocks.6.mlp.gate_proj",
	"visual.blocks.6.mlp.gate_proj.bias",
	"visual.blocks.6.mlp.up_proj",
	"visual.blocks.6.mlp.up_proj.bias",
	"visual.blocks.6.norm1",
	"visual.blocks.6.norm2",
	"visual.blocks.7.attn.k_norm",
	"visual.blocks.7.attn.proj",
	"visual.blocks.7.attn.proj.bias",
	"visual.blocks.7.attn.q_norm",
	"visual.blocks.7.attn.qkv",
	"visual.blocks.7.attn.qkv.bias",
	"visual.blocks.7.mlp.down_proj",
	"visual.blocks.7.mlp.down_proj.bias",
	"visual.blocks.7.mlp.gate_proj",
	"visual.blocks.7.mlp.gate_proj.bias",
	"visual.blocks.7.mlp.up_proj",
	"visual.blocks.7.mlp.up_proj.bias",
	"visual.blocks.7.norm1",
	"visual.blocks.7.norm2",
	"visual.blocks.8.attn.k_norm",
	"visual.blocks.8.attn.proj",
	"visual.blocks.8.attn.proj.bias",
	"visual.blocks.8.attn.q_norm",
	"visual.blocks.8.attn.qkv",
	"visual.blocks.8.attn.qkv.bias",
	"visual.blocks.8.mlp.down_proj",
	"visual.blocks.8.mlp.down_proj.bias",
	"visual.blocks.8.mlp.gate_proj",
	"visual.blocks.8.mlp.gate_proj.bias",
	"visual.blocks.8.mlp.up_proj",
	"visual.blocks.8.mlp.up_proj.bias",
	"visual.blocks.8.norm1",
	"visual.blocks.8.norm2",
	"visual.blocks.9.attn.k_norm",
	"visual.blocks.9.attn.proj",
	"visual.blocks.9.attn.proj.bias",
	"visual.blocks.9.attn.q_norm",
	"visual.blocks.9.attn.qkv",
	"visual.blocks.9.attn.qkv.bias",
	"visual.blocks.9.mlp.down_proj",
	"visual.blocks.9.mlp.down_proj.bias",
	"visual.blocks.9.mlp.gate_proj",
	"visual.blocks.9.mlp.gate_proj.bias",
	"visual.blocks.9.mlp.up_proj",
	"visual.blocks.9.mlp.up_proj.bias",
	"visual.blocks.9.norm1",
	"visual.blocks.9.norm2",
	"visual.downsample",
	"visual.downsample.bias",
	"visual.merger.down_proj",
	"visual.merger.gate_proj",
	"visual.merger.post_projection_norm",
	"visual.merger.post_projection_norm.bias",
	"visual.merger.proj",
	"visual.merger.up_proj",
	"visual.patch_embed.proj",
	"visual.patch_embed.proj.bias",
	"visual.post_layernorm",
	"weights_proj"
	],
	"quant_method": "fp8",
	"weight_block_size": [
	128,
	128
	]
	}
	}