ljoana's picture
Upload folder using huggingface_hub
f1dfc9e verified
{
"add_cross_attention": false,
"architectures": [
"Qwen2_5_VLForConditionalGeneration"
],
"attention_dropout": 0.0,
"bad_words_ids": null,
"begin_suppress_tokens": null,
"bos_token_id": 151643,
"chunk_size_feed_forward": 0,
"cross_attention_hidden_size": null,
"decoder_start_token_id": null,
"diversity_penalty": 0.0,
"do_sample": false,
"early_stopping": false,
"encoder_no_repeat_ngram_size": 0,
"eos_token_id": 151645,
"exponential_decay_length_penalty": null,
"finetuning_task": null,
"forced_bos_token_id": null,
"forced_eos_token_id": null,
"hidden_act": "silu",
"hidden_size": 3584,
"id2label": {
"0": "LABEL_0",
"1": "LABEL_1"
},
"image_token_id": 151655,
"initializer_range": 0.02,
"intermediate_size": 18944,
"is_decoder": false,
"is_encoder_decoder": false,
"label2id": {
"LABEL_0": 0,
"LABEL_1": 1
},
"length_penalty": 1.0,
"max_length": 20,
"max_position_embeddings": 128000,
"max_window_layers": 28,
"min_length": 0,
"model_type": "qwen2_5_vl",
"no_repeat_ngram_size": 0,
"num_attention_heads": 28,
"num_beam_groups": 1,
"num_beams": 1,
"num_hidden_layers": 28,
"num_key_value_heads": 4,
"num_return_sequences": 1,
"output_attentions": false,
"output_hidden_states": false,
"output_scores": false,
"pad_token_id": null,
"prefix": null,
"problem_type": null,
"pruned_heads": {},
"quantization": {
"group_size": 64,
"bits": 4,
"vision_tower.patch_embed.proj": false,
"vision_tower.rotary_pos_emb": false,
"vision_tower.blocks.0.norm1": false,
"vision_tower.blocks.0.norm2": false,
"vision_tower.blocks.0.attn.qkv": false,
"vision_tower.blocks.0.attn.proj": false,
"vision_tower.blocks.0.mlp.gate_proj": false,
"vision_tower.blocks.0.mlp.up_proj": false,
"vision_tower.blocks.0.mlp.down_proj": false,
"vision_tower.blocks.1.norm1": false,
"vision_tower.blocks.1.norm2": false,
"vision_tower.blocks.1.attn.qkv": false,
"vision_tower.blocks.1.attn.proj": false,
"vision_tower.blocks.1.mlp.gate_proj": false,
"vision_tower.blocks.1.mlp.up_proj": false,
"vision_tower.blocks.1.mlp.down_proj": false,
"vision_tower.blocks.2.norm1": false,
"vision_tower.blocks.2.norm2": false,
"vision_tower.blocks.2.attn.qkv": false,
"vision_tower.blocks.2.attn.proj": false,
"vision_tower.blocks.2.mlp.gate_proj": false,
"vision_tower.blocks.2.mlp.up_proj": false,
"vision_tower.blocks.2.mlp.down_proj": false,
"vision_tower.blocks.3.norm1": false,
"vision_tower.blocks.3.norm2": false,
"vision_tower.blocks.3.attn.qkv": false,
"vision_tower.blocks.3.attn.proj": false,
"vision_tower.blocks.3.mlp.gate_proj": false,
"vision_tower.blocks.3.mlp.up_proj": false,
"vision_tower.blocks.3.mlp.down_proj": false,
"vision_tower.blocks.4.norm1": false,
"vision_tower.blocks.4.norm2": false,
"vision_tower.blocks.4.attn.qkv": false,
"vision_tower.blocks.4.attn.proj": false,
"vision_tower.blocks.4.mlp.gate_proj": false,
"vision_tower.blocks.4.mlp.up_proj": false,
"vision_tower.blocks.4.mlp.down_proj": false,
"vision_tower.blocks.5.norm1": false,
"vision_tower.blocks.5.norm2": false,
"vision_tower.blocks.5.attn.qkv": false,
"vision_tower.blocks.5.attn.proj": false,
"vision_tower.blocks.5.mlp.gate_proj": false,
"vision_tower.blocks.5.mlp.up_proj": false,
"vision_tower.blocks.5.mlp.down_proj": false,
"vision_tower.blocks.6.norm1": false,
"vision_tower.blocks.6.norm2": false,
"vision_tower.blocks.6.attn.qkv": false,
"vision_tower.blocks.6.attn.proj": false,
"vision_tower.blocks.6.mlp.gate_proj": false,
"vision_tower.blocks.6.mlp.up_proj": false,
"vision_tower.blocks.6.mlp.down_proj": false,
"vision_tower.blocks.7.norm1": false,
"vision_tower.blocks.7.norm2": false,
"vision_tower.blocks.7.attn.qkv": false,
"vision_tower.blocks.7.attn.proj": false,
"vision_tower.blocks.7.mlp.gate_proj": false,
"vision_tower.blocks.7.mlp.up_proj": false,
"vision_tower.blocks.7.mlp.down_proj": false,
"vision_tower.blocks.8.norm1": false,
"vision_tower.blocks.8.norm2": false,
"vision_tower.blocks.8.attn.qkv": false,
"vision_tower.blocks.8.attn.proj": false,
"vision_tower.blocks.8.mlp.gate_proj": false,
"vision_tower.blocks.8.mlp.up_proj": false,
"vision_tower.blocks.8.mlp.down_proj": false,
"vision_tower.blocks.9.norm1": false,
"vision_tower.blocks.9.norm2": false,
"vision_tower.blocks.9.attn.qkv": false,
"vision_tower.blocks.9.attn.proj": false,
"vision_tower.blocks.9.mlp.gate_proj": false,
"vision_tower.blocks.9.mlp.up_proj": false,
"vision_tower.blocks.9.mlp.down_proj": false,
"vision_tower.blocks.10.norm1": false,
"vision_tower.blocks.10.norm2": false,
"vision_tower.blocks.10.attn.qkv": false,
"vision_tower.blocks.10.attn.proj": false,
"vision_tower.blocks.10.mlp.gate_proj": false,
"vision_tower.blocks.10.mlp.up_proj": false,
"vision_tower.blocks.10.mlp.down_proj": false,
"vision_tower.blocks.11.norm1": false,
"vision_tower.blocks.11.norm2": false,
"vision_tower.blocks.11.attn.qkv": false,
"vision_tower.blocks.11.attn.proj": false,
"vision_tower.blocks.11.mlp.gate_proj": false,
"vision_tower.blocks.11.mlp.up_proj": false,
"vision_tower.blocks.11.mlp.down_proj": false,
"vision_tower.blocks.12.norm1": false,
"vision_tower.blocks.12.norm2": false,
"vision_tower.blocks.12.attn.qkv": false,
"vision_tower.blocks.12.attn.proj": false,
"vision_tower.blocks.12.mlp.gate_proj": false,
"vision_tower.blocks.12.mlp.up_proj": false,
"vision_tower.blocks.12.mlp.down_proj": false,
"vision_tower.blocks.13.norm1": false,
"vision_tower.blocks.13.norm2": false,
"vision_tower.blocks.13.attn.qkv": false,
"vision_tower.blocks.13.attn.proj": false,
"vision_tower.blocks.13.mlp.gate_proj": false,
"vision_tower.blocks.13.mlp.up_proj": false,
"vision_tower.blocks.13.mlp.down_proj": false,
"vision_tower.blocks.14.norm1": false,
"vision_tower.blocks.14.norm2": false,
"vision_tower.blocks.14.attn.qkv": false,
"vision_tower.blocks.14.attn.proj": false,
"vision_tower.blocks.14.mlp.gate_proj": false,
"vision_tower.blocks.14.mlp.up_proj": false,
"vision_tower.blocks.14.mlp.down_proj": false,
"vision_tower.blocks.15.norm1": false,
"vision_tower.blocks.15.norm2": false,
"vision_tower.blocks.15.attn.qkv": false,
"vision_tower.blocks.15.attn.proj": false,
"vision_tower.blocks.15.mlp.gate_proj": false,
"vision_tower.blocks.15.mlp.up_proj": false,
"vision_tower.blocks.15.mlp.down_proj": false,
"vision_tower.blocks.16.norm1": false,
"vision_tower.blocks.16.norm2": false,
"vision_tower.blocks.16.attn.qkv": false,
"vision_tower.blocks.16.attn.proj": false,
"vision_tower.blocks.16.mlp.gate_proj": false,
"vision_tower.blocks.16.mlp.up_proj": false,
"vision_tower.blocks.16.mlp.down_proj": false,
"vision_tower.blocks.17.norm1": false,
"vision_tower.blocks.17.norm2": false,
"vision_tower.blocks.17.attn.qkv": false,
"vision_tower.blocks.17.attn.proj": false,
"vision_tower.blocks.17.mlp.gate_proj": false,
"vision_tower.blocks.17.mlp.up_proj": false,
"vision_tower.blocks.17.mlp.down_proj": false,
"vision_tower.blocks.18.norm1": false,
"vision_tower.blocks.18.norm2": false,
"vision_tower.blocks.18.attn.qkv": false,
"vision_tower.blocks.18.attn.proj": false,
"vision_tower.blocks.18.mlp.gate_proj": false,
"vision_tower.blocks.18.mlp.up_proj": false,
"vision_tower.blocks.18.mlp.down_proj": false,
"vision_tower.blocks.19.norm1": false,
"vision_tower.blocks.19.norm2": false,
"vision_tower.blocks.19.attn.qkv": false,
"vision_tower.blocks.19.attn.proj": false,
"vision_tower.blocks.19.mlp.gate_proj": false,
"vision_tower.blocks.19.mlp.up_proj": false,
"vision_tower.blocks.19.mlp.down_proj": false,
"vision_tower.blocks.20.norm1": false,
"vision_tower.blocks.20.norm2": false,
"vision_tower.blocks.20.attn.qkv": false,
"vision_tower.blocks.20.attn.proj": false,
"vision_tower.blocks.20.mlp.gate_proj": false,
"vision_tower.blocks.20.mlp.up_proj": false,
"vision_tower.blocks.20.mlp.down_proj": false,
"vision_tower.blocks.21.norm1": false,
"vision_tower.blocks.21.norm2": false,
"vision_tower.blocks.21.attn.qkv": false,
"vision_tower.blocks.21.attn.proj": false,
"vision_tower.blocks.21.mlp.gate_proj": false,
"vision_tower.blocks.21.mlp.up_proj": false,
"vision_tower.blocks.21.mlp.down_proj": false,
"vision_tower.blocks.22.norm1": false,
"vision_tower.blocks.22.norm2": false,
"vision_tower.blocks.22.attn.qkv": false,
"vision_tower.blocks.22.attn.proj": false,
"vision_tower.blocks.22.mlp.gate_proj": false,
"vision_tower.blocks.22.mlp.up_proj": false,
"vision_tower.blocks.22.mlp.down_proj": false,
"vision_tower.blocks.23.norm1": false,
"vision_tower.blocks.23.norm2": false,
"vision_tower.blocks.23.attn.qkv": false,
"vision_tower.blocks.23.attn.proj": false,
"vision_tower.blocks.23.mlp.gate_proj": false,
"vision_tower.blocks.23.mlp.up_proj": false,
"vision_tower.blocks.23.mlp.down_proj": false,
"vision_tower.blocks.24.norm1": false,
"vision_tower.blocks.24.norm2": false,
"vision_tower.blocks.24.attn.qkv": false,
"vision_tower.blocks.24.attn.proj": false,
"vision_tower.blocks.24.mlp.gate_proj": false,
"vision_tower.blocks.24.mlp.up_proj": false,
"vision_tower.blocks.24.mlp.down_proj": false,
"vision_tower.blocks.25.norm1": false,
"vision_tower.blocks.25.norm2": false,
"vision_tower.blocks.25.attn.qkv": false,
"vision_tower.blocks.25.attn.proj": false,
"vision_tower.blocks.25.mlp.gate_proj": false,
"vision_tower.blocks.25.mlp.up_proj": false,
"vision_tower.blocks.25.mlp.down_proj": false,
"vision_tower.blocks.26.norm1": false,
"vision_tower.blocks.26.norm2": false,
"vision_tower.blocks.26.attn.qkv": false,
"vision_tower.blocks.26.attn.proj": false,
"vision_tower.blocks.26.mlp.gate_proj": false,
"vision_tower.blocks.26.mlp.up_proj": false,
"vision_tower.blocks.26.mlp.down_proj": false,
"vision_tower.blocks.27.norm1": false,
"vision_tower.blocks.27.norm2": false,
"vision_tower.blocks.27.attn.qkv": false,
"vision_tower.blocks.27.attn.proj": false,
"vision_tower.blocks.27.mlp.gate_proj": false,
"vision_tower.blocks.27.mlp.up_proj": false,
"vision_tower.blocks.27.mlp.down_proj": false,
"vision_tower.blocks.28.norm1": false,
"vision_tower.blocks.28.norm2": false,
"vision_tower.blocks.28.attn.qkv": false,
"vision_tower.blocks.28.attn.proj": false,
"vision_tower.blocks.28.mlp.gate_proj": false,
"vision_tower.blocks.28.mlp.up_proj": false,
"vision_tower.blocks.28.mlp.down_proj": false,
"vision_tower.blocks.29.norm1": false,
"vision_tower.blocks.29.norm2": false,
"vision_tower.blocks.29.attn.qkv": false,
"vision_tower.blocks.29.attn.proj": false,
"vision_tower.blocks.29.mlp.gate_proj": false,
"vision_tower.blocks.29.mlp.up_proj": false,
"vision_tower.blocks.29.mlp.down_proj": false,
"vision_tower.blocks.30.norm1": false,
"vision_tower.blocks.30.norm2": false,
"vision_tower.blocks.30.attn.qkv": false,
"vision_tower.blocks.30.attn.proj": false,
"vision_tower.blocks.30.mlp.gate_proj": false,
"vision_tower.blocks.30.mlp.up_proj": false,
"vision_tower.blocks.30.mlp.down_proj": false,
"vision_tower.blocks.31.norm1": false,
"vision_tower.blocks.31.norm2": false,
"vision_tower.blocks.31.attn.qkv": false,
"vision_tower.blocks.31.attn.proj": false,
"vision_tower.blocks.31.mlp.gate_proj": false,
"vision_tower.blocks.31.mlp.up_proj": false,
"vision_tower.blocks.31.mlp.down_proj": false,
"vision_tower.merger.ln_q": false,
"vision_tower.merger.mlp.0": false,
"vision_tower.merger.mlp.1": false,
"vision_tower.merger.mlp.2": false,
"language_model.model.embed_tokens": true,
"language_model.model.layers.0.self_attn.q_proj": true,
"language_model.model.layers.0.self_attn.k_proj": true,
"language_model.model.layers.0.self_attn.v_proj": true,
"language_model.model.layers.0.self_attn.o_proj": true,
"language_model.model.layers.0.mlp.gate_proj": true,
"language_model.model.layers.0.mlp.down_proj": true,
"language_model.model.layers.0.mlp.up_proj": true,
"language_model.model.layers.0.input_layernorm": false,
"language_model.model.layers.0.post_attention_layernorm": false,
"language_model.model.layers.1.self_attn.q_proj": true,
"language_model.model.layers.1.self_attn.k_proj": true,
"language_model.model.layers.1.self_attn.v_proj": true,
"language_model.model.layers.1.self_attn.o_proj": true,
"language_model.model.layers.1.mlp.gate_proj": true,
"language_model.model.layers.1.mlp.down_proj": true,
"language_model.model.layers.1.mlp.up_proj": true,
"language_model.model.layers.1.input_layernorm": false,
"language_model.model.layers.1.post_attention_layernorm": false,
"language_model.model.layers.2.self_attn.q_proj": true,
"language_model.model.layers.2.self_attn.k_proj": true,
"language_model.model.layers.2.self_attn.v_proj": true,
"language_model.model.layers.2.self_attn.o_proj": true,
"language_model.model.layers.2.mlp.gate_proj": true,
"language_model.model.layers.2.mlp.down_proj": true,
"language_model.model.layers.2.mlp.up_proj": true,
"language_model.model.layers.2.input_layernorm": false,
"language_model.model.layers.2.post_attention_layernorm": false,
"language_model.model.layers.3.self_attn.q_proj": true,
"language_model.model.layers.3.self_attn.k_proj": true,
"language_model.model.layers.3.self_attn.v_proj": true,
"language_model.model.layers.3.self_attn.o_proj": true,
"language_model.model.layers.3.mlp.gate_proj": true,
"language_model.model.layers.3.mlp.down_proj": true,
"language_model.model.layers.3.mlp.up_proj": true,
"language_model.model.layers.3.input_layernorm": false,
"language_model.model.layers.3.post_attention_layernorm": false,
"language_model.model.layers.4.self_attn.q_proj": true,
"language_model.model.layers.4.self_attn.k_proj": true,
"language_model.model.layers.4.self_attn.v_proj": true,
"language_model.model.layers.4.self_attn.o_proj": true,
"language_model.model.layers.4.mlp.gate_proj": true,
"language_model.model.layers.4.mlp.down_proj": true,
"language_model.model.layers.4.mlp.up_proj": true,
"language_model.model.layers.4.input_layernorm": false,
"language_model.model.layers.4.post_attention_layernorm": false,
"language_model.model.layers.5.self_attn.q_proj": true,
"language_model.model.layers.5.self_attn.k_proj": true,
"language_model.model.layers.5.self_attn.v_proj": true,
"language_model.model.layers.5.self_attn.o_proj": true,
"language_model.model.layers.5.mlp.gate_proj": true,
"language_model.model.layers.5.mlp.down_proj": true,
"language_model.model.layers.5.mlp.up_proj": true,
"language_model.model.layers.5.input_layernorm": false,
"language_model.model.layers.5.post_attention_layernorm": false,
"language_model.model.layers.6.self_attn.q_proj": true,
"language_model.model.layers.6.self_attn.k_proj": true,
"language_model.model.layers.6.self_attn.v_proj": true,
"language_model.model.layers.6.self_attn.o_proj": true,
"language_model.model.layers.6.mlp.gate_proj": true,
"language_model.model.layers.6.mlp.down_proj": true,
"language_model.model.layers.6.mlp.up_proj": true,
"language_model.model.layers.6.input_layernorm": false,
"language_model.model.layers.6.post_attention_layernorm": false,
"language_model.model.layers.7.self_attn.q_proj": true,
"language_model.model.layers.7.self_attn.k_proj": true,
"language_model.model.layers.7.self_attn.v_proj": true,
"language_model.model.layers.7.self_attn.o_proj": true,
"language_model.model.layers.7.mlp.gate_proj": true,
"language_model.model.layers.7.mlp.down_proj": true,
"language_model.model.layers.7.mlp.up_proj": true,
"language_model.model.layers.7.input_layernorm": false,
"language_model.model.layers.7.post_attention_layernorm": false,
"language_model.model.layers.8.self_attn.q_proj": true,
"language_model.model.layers.8.self_attn.k_proj": true,
"language_model.model.layers.8.self_attn.v_proj": true,
"language_model.model.layers.8.self_attn.o_proj": true,
"language_model.model.layers.8.mlp.gate_proj": true,
"language_model.model.layers.8.mlp.down_proj": true,
"language_model.model.layers.8.mlp.up_proj": true,
"language_model.model.layers.8.input_layernorm": false,
"language_model.model.layers.8.post_attention_layernorm": false,
"language_model.model.layers.9.self_attn.q_proj": true,
"language_model.model.layers.9.self_attn.k_proj": true,
"language_model.model.layers.9.self_attn.v_proj": true,
"language_model.model.layers.9.self_attn.o_proj": true,
"language_model.model.layers.9.mlp.gate_proj": true,
"language_model.model.layers.9.mlp.down_proj": true,
"language_model.model.layers.9.mlp.up_proj": true,
"language_model.model.layers.9.input_layernorm": false,
"language_model.model.layers.9.post_attention_layernorm": false,
"language_model.model.layers.10.self_attn.q_proj": true,
"language_model.model.layers.10.self_attn.k_proj": true,
"language_model.model.layers.10.self_attn.v_proj": true,
"language_model.model.layers.10.self_attn.o_proj": true,
"language_model.model.layers.10.mlp.gate_proj": true,
"language_model.model.layers.10.mlp.down_proj": true,
"language_model.model.layers.10.mlp.up_proj": true,
"language_model.model.layers.10.input_layernorm": false,
"language_model.model.layers.10.post_attention_layernorm": false,
"language_model.model.layers.11.self_attn.q_proj": true,
"language_model.model.layers.11.self_attn.k_proj": true,
"language_model.model.layers.11.self_attn.v_proj": true,
"language_model.model.layers.11.self_attn.o_proj": true,
"language_model.model.layers.11.mlp.gate_proj": true,
"language_model.model.layers.11.mlp.down_proj": true,
"language_model.model.layers.11.mlp.up_proj": true,
"language_model.model.layers.11.input_layernorm": false,
"language_model.model.layers.11.post_attention_layernorm": false,
"language_model.model.layers.12.self_attn.q_proj": true,
"language_model.model.layers.12.self_attn.k_proj": true,
"language_model.model.layers.12.self_attn.v_proj": true,
"language_model.model.layers.12.self_attn.o_proj": true,
"language_model.model.layers.12.mlp.gate_proj": true,
"language_model.model.layers.12.mlp.down_proj": true,
"language_model.model.layers.12.mlp.up_proj": true,
"language_model.model.layers.12.input_layernorm": false,
"language_model.model.layers.12.post_attention_layernorm": false,
"language_model.model.layers.13.self_attn.q_proj": true,
"language_model.model.layers.13.self_attn.k_proj": true,
"language_model.model.layers.13.self_attn.v_proj": true,
"language_model.model.layers.13.self_attn.o_proj": true,
"language_model.model.layers.13.mlp.gate_proj": true,
"language_model.model.layers.13.mlp.down_proj": true,
"language_model.model.layers.13.mlp.up_proj": true,
"language_model.model.layers.13.input_layernorm": false,
"language_model.model.layers.13.post_attention_layernorm": false,
"language_model.model.layers.14.self_attn.q_proj": true,
"language_model.model.layers.14.self_attn.k_proj": true,
"language_model.model.layers.14.self_attn.v_proj": true,
"language_model.model.layers.14.self_attn.o_proj": true,
"language_model.model.layers.14.mlp.gate_proj": true,
"language_model.model.layers.14.mlp.down_proj": true,
"language_model.model.layers.14.mlp.up_proj": true,
"language_model.model.layers.14.input_layernorm": false,
"language_model.model.layers.14.post_attention_layernorm": false,
"language_model.model.layers.15.self_attn.q_proj": true,
"language_model.model.layers.15.self_attn.k_proj": true,
"language_model.model.layers.15.self_attn.v_proj": true,
"language_model.model.layers.15.self_attn.o_proj": true,
"language_model.model.layers.15.mlp.gate_proj": true,
"language_model.model.layers.15.mlp.down_proj": true,
"language_model.model.layers.15.mlp.up_proj": true,
"language_model.model.layers.15.input_layernorm": false,
"language_model.model.layers.15.post_attention_layernorm": false,
"language_model.model.layers.16.self_attn.q_proj": true,
"language_model.model.layers.16.self_attn.k_proj": true,
"language_model.model.layers.16.self_attn.v_proj": true,
"language_model.model.layers.16.self_attn.o_proj": true,
"language_model.model.layers.16.mlp.gate_proj": true,
"language_model.model.layers.16.mlp.down_proj": true,
"language_model.model.layers.16.mlp.up_proj": true,
"language_model.model.layers.16.input_layernorm": false,
"language_model.model.layers.16.post_attention_layernorm": false,
"language_model.model.layers.17.self_attn.q_proj": true,
"language_model.model.layers.17.self_attn.k_proj": true,
"language_model.model.layers.17.self_attn.v_proj": true,
"language_model.model.layers.17.self_attn.o_proj": true,
"language_model.model.layers.17.mlp.gate_proj": true,
"language_model.model.layers.17.mlp.down_proj": true,
"language_model.model.layers.17.mlp.up_proj": true,
"language_model.model.layers.17.input_layernorm": false,
"language_model.model.layers.17.post_attention_layernorm": false,
"language_model.model.layers.18.self_attn.q_proj": true,
"language_model.model.layers.18.self_attn.k_proj": true,
"language_model.model.layers.18.self_attn.v_proj": true,
"language_model.model.layers.18.self_attn.o_proj": true,
"language_model.model.layers.18.mlp.gate_proj": true,
"language_model.model.layers.18.mlp.down_proj": true,
"language_model.model.layers.18.mlp.up_proj": true,
"language_model.model.layers.18.input_layernorm": false,
"language_model.model.layers.18.post_attention_layernorm": false,
"language_model.model.layers.19.self_attn.q_proj": true,
"language_model.model.layers.19.self_attn.k_proj": true,
"language_model.model.layers.19.self_attn.v_proj": true,
"language_model.model.layers.19.self_attn.o_proj": true,
"language_model.model.layers.19.mlp.gate_proj": true,
"language_model.model.layers.19.mlp.down_proj": true,
"language_model.model.layers.19.mlp.up_proj": true,
"language_model.model.layers.19.input_layernorm": false,
"language_model.model.layers.19.post_attention_layernorm": false,
"language_model.model.layers.20.self_attn.q_proj": true,
"language_model.model.layers.20.self_attn.k_proj": true,
"language_model.model.layers.20.self_attn.v_proj": true,
"language_model.model.layers.20.self_attn.o_proj": true,
"language_model.model.layers.20.mlp.gate_proj": true,
"language_model.model.layers.20.mlp.down_proj": true,
"language_model.model.layers.20.mlp.up_proj": true,
"language_model.model.layers.20.input_layernorm": false,
"language_model.model.layers.20.post_attention_layernorm": false,
"language_model.model.layers.21.self_attn.q_proj": true,
"language_model.model.layers.21.self_attn.k_proj": true,
"language_model.model.layers.21.self_attn.v_proj": true,
"language_model.model.layers.21.self_attn.o_proj": true,
"language_model.model.layers.21.mlp.gate_proj": true,
"language_model.model.layers.21.mlp.down_proj": true,
"language_model.model.layers.21.mlp.up_proj": true,
"language_model.model.layers.21.input_layernorm": false,
"language_model.model.layers.21.post_attention_layernorm": false,
"language_model.model.layers.22.self_attn.q_proj": true,
"language_model.model.layers.22.self_attn.k_proj": true,
"language_model.model.layers.22.self_attn.v_proj": true,
"language_model.model.layers.22.self_attn.o_proj": true,
"language_model.model.layers.22.mlp.gate_proj": true,
"language_model.model.layers.22.mlp.down_proj": true,
"language_model.model.layers.22.mlp.up_proj": true,
"language_model.model.layers.22.input_layernorm": false,
"language_model.model.layers.22.post_attention_layernorm": false,
"language_model.model.layers.23.self_attn.q_proj": true,
"language_model.model.layers.23.self_attn.k_proj": true,
"language_model.model.layers.23.self_attn.v_proj": true,
"language_model.model.layers.23.self_attn.o_proj": true,
"language_model.model.layers.23.mlp.gate_proj": true,
"language_model.model.layers.23.mlp.down_proj": true,
"language_model.model.layers.23.mlp.up_proj": true,
"language_model.model.layers.23.input_layernorm": false,
"language_model.model.layers.23.post_attention_layernorm": false,
"language_model.model.layers.24.self_attn.q_proj": true,
"language_model.model.layers.24.self_attn.k_proj": true,
"language_model.model.layers.24.self_attn.v_proj": true,
"language_model.model.layers.24.self_attn.o_proj": true,
"language_model.model.layers.24.mlp.gate_proj": true,
"language_model.model.layers.24.mlp.down_proj": true,
"language_model.model.layers.24.mlp.up_proj": true,
"language_model.model.layers.24.input_layernorm": false,
"language_model.model.layers.24.post_attention_layernorm": false,
"language_model.model.layers.25.self_attn.q_proj": true,
"language_model.model.layers.25.self_attn.k_proj": true,
"language_model.model.layers.25.self_attn.v_proj": true,
"language_model.model.layers.25.self_attn.o_proj": true,
"language_model.model.layers.25.mlp.gate_proj": true,
"language_model.model.layers.25.mlp.down_proj": true,
"language_model.model.layers.25.mlp.up_proj": true,
"language_model.model.layers.25.input_layernorm": false,
"language_model.model.layers.25.post_attention_layernorm": false,
"language_model.model.layers.26.self_attn.q_proj": true,
"language_model.model.layers.26.self_attn.k_proj": true,
"language_model.model.layers.26.self_attn.v_proj": true,
"language_model.model.layers.26.self_attn.o_proj": true,
"language_model.model.layers.26.mlp.gate_proj": true,
"language_model.model.layers.26.mlp.down_proj": true,
"language_model.model.layers.26.mlp.up_proj": true,
"language_model.model.layers.26.input_layernorm": false,
"language_model.model.layers.26.post_attention_layernorm": false,
"language_model.model.layers.27.self_attn.q_proj": true,
"language_model.model.layers.27.self_attn.k_proj": true,
"language_model.model.layers.27.self_attn.v_proj": true,
"language_model.model.layers.27.self_attn.o_proj": true,
"language_model.model.layers.27.mlp.gate_proj": true,
"language_model.model.layers.27.mlp.down_proj": true,
"language_model.model.layers.27.mlp.up_proj": true,
"language_model.model.layers.27.input_layernorm": false,
"language_model.model.layers.27.post_attention_layernorm": false,
"language_model.model.norm": false,
"language_model.lm_head": true
},
"quantization_config": {
"group_size": 64,
"bits": 4,
"vision_tower.patch_embed.proj": false,
"vision_tower.rotary_pos_emb": false,
"vision_tower.blocks.0.norm1": false,
"vision_tower.blocks.0.norm2": false,
"vision_tower.blocks.0.attn.qkv": false,
"vision_tower.blocks.0.attn.proj": false,
"vision_tower.blocks.0.mlp.gate_proj": false,
"vision_tower.blocks.0.mlp.up_proj": false,
"vision_tower.blocks.0.mlp.down_proj": false,
"vision_tower.blocks.1.norm1": false,
"vision_tower.blocks.1.norm2": false,
"vision_tower.blocks.1.attn.qkv": false,
"vision_tower.blocks.1.attn.proj": false,
"vision_tower.blocks.1.mlp.gate_proj": false,
"vision_tower.blocks.1.mlp.up_proj": false,
"vision_tower.blocks.1.mlp.down_proj": false,
"vision_tower.blocks.2.norm1": false,
"vision_tower.blocks.2.norm2": false,
"vision_tower.blocks.2.attn.qkv": false,
"vision_tower.blocks.2.attn.proj": false,
"vision_tower.blocks.2.mlp.gate_proj": false,
"vision_tower.blocks.2.mlp.up_proj": false,
"vision_tower.blocks.2.mlp.down_proj": false,
"vision_tower.blocks.3.norm1": false,
"vision_tower.blocks.3.norm2": false,
"vision_tower.blocks.3.attn.qkv": false,
"vision_tower.blocks.3.attn.proj": false,
"vision_tower.blocks.3.mlp.gate_proj": false,
"vision_tower.blocks.3.mlp.up_proj": false,
"vision_tower.blocks.3.mlp.down_proj": false,
"vision_tower.blocks.4.norm1": false,
"vision_tower.blocks.4.norm2": false,
"vision_tower.blocks.4.attn.qkv": false,
"vision_tower.blocks.4.attn.proj": false,
"vision_tower.blocks.4.mlp.gate_proj": false,
"vision_tower.blocks.4.mlp.up_proj": false,
"vision_tower.blocks.4.mlp.down_proj": false,
"vision_tower.blocks.5.norm1": false,
"vision_tower.blocks.5.norm2": false,
"vision_tower.blocks.5.attn.qkv": false,
"vision_tower.blocks.5.attn.proj": false,
"vision_tower.blocks.5.mlp.gate_proj": false,
"vision_tower.blocks.5.mlp.up_proj": false,
"vision_tower.blocks.5.mlp.down_proj": false,
"vision_tower.blocks.6.norm1": false,
"vision_tower.blocks.6.norm2": false,
"vision_tower.blocks.6.attn.qkv": false,
"vision_tower.blocks.6.attn.proj": false,
"vision_tower.blocks.6.mlp.gate_proj": false,
"vision_tower.blocks.6.mlp.up_proj": false,
"vision_tower.blocks.6.mlp.down_proj": false,
"vision_tower.blocks.7.norm1": false,
"vision_tower.blocks.7.norm2": false,
"vision_tower.blocks.7.attn.qkv": false,
"vision_tower.blocks.7.attn.proj": false,
"vision_tower.blocks.7.mlp.gate_proj": false,
"vision_tower.blocks.7.mlp.up_proj": false,
"vision_tower.blocks.7.mlp.down_proj": false,
"vision_tower.blocks.8.norm1": false,
"vision_tower.blocks.8.norm2": false,
"vision_tower.blocks.8.attn.qkv": false,
"vision_tower.blocks.8.attn.proj": false,
"vision_tower.blocks.8.mlp.gate_proj": false,
"vision_tower.blocks.8.mlp.up_proj": false,
"vision_tower.blocks.8.mlp.down_proj": false,
"vision_tower.blocks.9.norm1": false,
"vision_tower.blocks.9.norm2": false,
"vision_tower.blocks.9.attn.qkv": false,
"vision_tower.blocks.9.attn.proj": false,
"vision_tower.blocks.9.mlp.gate_proj": false,
"vision_tower.blocks.9.mlp.up_proj": false,
"vision_tower.blocks.9.mlp.down_proj": false,
"vision_tower.blocks.10.norm1": false,
"vision_tower.blocks.10.norm2": false,
"vision_tower.blocks.10.attn.qkv": false,
"vision_tower.blocks.10.attn.proj": false,
"vision_tower.blocks.10.mlp.gate_proj": false,
"vision_tower.blocks.10.mlp.up_proj": false,
"vision_tower.blocks.10.mlp.down_proj": false,
"vision_tower.blocks.11.norm1": false,
"vision_tower.blocks.11.norm2": false,
"vision_tower.blocks.11.attn.qkv": false,
"vision_tower.blocks.11.attn.proj": false,
"vision_tower.blocks.11.mlp.gate_proj": false,
"vision_tower.blocks.11.mlp.up_proj": false,
"vision_tower.blocks.11.mlp.down_proj": false,
"vision_tower.blocks.12.norm1": false,
"vision_tower.blocks.12.norm2": false,
"vision_tower.blocks.12.attn.qkv": false,
"vision_tower.blocks.12.attn.proj": false,
"vision_tower.blocks.12.mlp.gate_proj": false,
"vision_tower.blocks.12.mlp.up_proj": false,
"vision_tower.blocks.12.mlp.down_proj": false,
"vision_tower.blocks.13.norm1": false,
"vision_tower.blocks.13.norm2": false,
"vision_tower.blocks.13.attn.qkv": false,
"vision_tower.blocks.13.attn.proj": false,
"vision_tower.blocks.13.mlp.gate_proj": false,
"vision_tower.blocks.13.mlp.up_proj": false,
"vision_tower.blocks.13.mlp.down_proj": false,
"vision_tower.blocks.14.norm1": false,
"vision_tower.blocks.14.norm2": false,
"vision_tower.blocks.14.attn.qkv": false,
"vision_tower.blocks.14.attn.proj": false,
"vision_tower.blocks.14.mlp.gate_proj": false,
"vision_tower.blocks.14.mlp.up_proj": false,
"vision_tower.blocks.14.mlp.down_proj": false,
"vision_tower.blocks.15.norm1": false,
"vision_tower.blocks.15.norm2": false,
"vision_tower.blocks.15.attn.qkv": false,
"vision_tower.blocks.15.attn.proj": false,
"vision_tower.blocks.15.mlp.gate_proj": false,
"vision_tower.blocks.15.mlp.up_proj": false,
"vision_tower.blocks.15.mlp.down_proj": false,
"vision_tower.blocks.16.norm1": false,
"vision_tower.blocks.16.norm2": false,
"vision_tower.blocks.16.attn.qkv": false,
"vision_tower.blocks.16.attn.proj": false,
"vision_tower.blocks.16.mlp.gate_proj": false,
"vision_tower.blocks.16.mlp.up_proj": false,
"vision_tower.blocks.16.mlp.down_proj": false,
"vision_tower.blocks.17.norm1": false,
"vision_tower.blocks.17.norm2": false,
"vision_tower.blocks.17.attn.qkv": false,
"vision_tower.blocks.17.attn.proj": false,
"vision_tower.blocks.17.mlp.gate_proj": false,
"vision_tower.blocks.17.mlp.up_proj": false,
"vision_tower.blocks.17.mlp.down_proj": false,
"vision_tower.blocks.18.norm1": false,
"vision_tower.blocks.18.norm2": false,
"vision_tower.blocks.18.attn.qkv": false,
"vision_tower.blocks.18.attn.proj": false,
"vision_tower.blocks.18.mlp.gate_proj": false,
"vision_tower.blocks.18.mlp.up_proj": false,
"vision_tower.blocks.18.mlp.down_proj": false,
"vision_tower.blocks.19.norm1": false,
"vision_tower.blocks.19.norm2": false,
"vision_tower.blocks.19.attn.qkv": false,
"vision_tower.blocks.19.attn.proj": false,
"vision_tower.blocks.19.mlp.gate_proj": false,
"vision_tower.blocks.19.mlp.up_proj": false,
"vision_tower.blocks.19.mlp.down_proj": false,
"vision_tower.blocks.20.norm1": false,
"vision_tower.blocks.20.norm2": false,
"vision_tower.blocks.20.attn.qkv": false,
"vision_tower.blocks.20.attn.proj": false,
"vision_tower.blocks.20.mlp.gate_proj": false,
"vision_tower.blocks.20.mlp.up_proj": false,
"vision_tower.blocks.20.mlp.down_proj": false,
"vision_tower.blocks.21.norm1": false,
"vision_tower.blocks.21.norm2": false,
"vision_tower.blocks.21.attn.qkv": false,
"vision_tower.blocks.21.attn.proj": false,
"vision_tower.blocks.21.mlp.gate_proj": false,
"vision_tower.blocks.21.mlp.up_proj": false,
"vision_tower.blocks.21.mlp.down_proj": false,
"vision_tower.blocks.22.norm1": false,
"vision_tower.blocks.22.norm2": false,
"vision_tower.blocks.22.attn.qkv": false,
"vision_tower.blocks.22.attn.proj": false,
"vision_tower.blocks.22.mlp.gate_proj": false,
"vision_tower.blocks.22.mlp.up_proj": false,
"vision_tower.blocks.22.mlp.down_proj": false,
"vision_tower.blocks.23.norm1": false,
"vision_tower.blocks.23.norm2": false,
"vision_tower.blocks.23.attn.qkv": false,
"vision_tower.blocks.23.attn.proj": false,
"vision_tower.blocks.23.mlp.gate_proj": false,
"vision_tower.blocks.23.mlp.up_proj": false,
"vision_tower.blocks.23.mlp.down_proj": false,
"vision_tower.blocks.24.norm1": false,
"vision_tower.blocks.24.norm2": false,
"vision_tower.blocks.24.attn.qkv": false,
"vision_tower.blocks.24.attn.proj": false,
"vision_tower.blocks.24.mlp.gate_proj": false,
"vision_tower.blocks.24.mlp.up_proj": false,
"vision_tower.blocks.24.mlp.down_proj": false,
"vision_tower.blocks.25.norm1": false,
"vision_tower.blocks.25.norm2": false,
"vision_tower.blocks.25.attn.qkv": false,
"vision_tower.blocks.25.attn.proj": false,
"vision_tower.blocks.25.mlp.gate_proj": false,
"vision_tower.blocks.25.mlp.up_proj": false,
"vision_tower.blocks.25.mlp.down_proj": false,
"vision_tower.blocks.26.norm1": false,
"vision_tower.blocks.26.norm2": false,
"vision_tower.blocks.26.attn.qkv": false,
"vision_tower.blocks.26.attn.proj": false,
"vision_tower.blocks.26.mlp.gate_proj": false,
"vision_tower.blocks.26.mlp.up_proj": false,
"vision_tower.blocks.26.mlp.down_proj": false,
"vision_tower.blocks.27.norm1": false,
"vision_tower.blocks.27.norm2": false,
"vision_tower.blocks.27.attn.qkv": false,
"vision_tower.blocks.27.attn.proj": false,
"vision_tower.blocks.27.mlp.gate_proj": false,
"vision_tower.blocks.27.mlp.up_proj": false,
"vision_tower.blocks.27.mlp.down_proj": false,
"vision_tower.blocks.28.norm1": false,
"vision_tower.blocks.28.norm2": false,
"vision_tower.blocks.28.attn.qkv": false,
"vision_tower.blocks.28.attn.proj": false,
"vision_tower.blocks.28.mlp.gate_proj": false,
"vision_tower.blocks.28.mlp.up_proj": false,
"vision_tower.blocks.28.mlp.down_proj": false,
"vision_tower.blocks.29.norm1": false,
"vision_tower.blocks.29.norm2": false,
"vision_tower.blocks.29.attn.qkv": false,
"vision_tower.blocks.29.attn.proj": false,
"vision_tower.blocks.29.mlp.gate_proj": false,
"vision_tower.blocks.29.mlp.up_proj": false,
"vision_tower.blocks.29.mlp.down_proj": false,
"vision_tower.blocks.30.norm1": false,
"vision_tower.blocks.30.norm2": false,
"vision_tower.blocks.30.attn.qkv": false,
"vision_tower.blocks.30.attn.proj": false,
"vision_tower.blocks.30.mlp.gate_proj": false,
"vision_tower.blocks.30.mlp.up_proj": false,
"vision_tower.blocks.30.mlp.down_proj": false,
"vision_tower.blocks.31.norm1": false,
"vision_tower.blocks.31.norm2": false,
"vision_tower.blocks.31.attn.qkv": false,
"vision_tower.blocks.31.attn.proj": false,
"vision_tower.blocks.31.mlp.gate_proj": false,
"vision_tower.blocks.31.mlp.up_proj": false,
"vision_tower.blocks.31.mlp.down_proj": false,
"vision_tower.merger.ln_q": false,
"vision_tower.merger.mlp.0": false,
"vision_tower.merger.mlp.1": false,
"vision_tower.merger.mlp.2": false,
"language_model.model.embed_tokens": true,
"language_model.model.layers.0.self_attn.q_proj": true,
"language_model.model.layers.0.self_attn.k_proj": true,
"language_model.model.layers.0.self_attn.v_proj": true,
"language_model.model.layers.0.self_attn.o_proj": true,
"language_model.model.layers.0.mlp.gate_proj": true,
"language_model.model.layers.0.mlp.down_proj": true,
"language_model.model.layers.0.mlp.up_proj": true,
"language_model.model.layers.0.input_layernorm": false,
"language_model.model.layers.0.post_attention_layernorm": false,
"language_model.model.layers.1.self_attn.q_proj": true,
"language_model.model.layers.1.self_attn.k_proj": true,
"language_model.model.layers.1.self_attn.v_proj": true,
"language_model.model.layers.1.self_attn.o_proj": true,
"language_model.model.layers.1.mlp.gate_proj": true,
"language_model.model.layers.1.mlp.down_proj": true,
"language_model.model.layers.1.mlp.up_proj": true,
"language_model.model.layers.1.input_layernorm": false,
"language_model.model.layers.1.post_attention_layernorm": false,
"language_model.model.layers.2.self_attn.q_proj": true,
"language_model.model.layers.2.self_attn.k_proj": true,
"language_model.model.layers.2.self_attn.v_proj": true,
"language_model.model.layers.2.self_attn.o_proj": true,
"language_model.model.layers.2.mlp.gate_proj": true,
"language_model.model.layers.2.mlp.down_proj": true,
"language_model.model.layers.2.mlp.up_proj": true,
"language_model.model.layers.2.input_layernorm": false,
"language_model.model.layers.2.post_attention_layernorm": false,
"language_model.model.layers.3.self_attn.q_proj": true,
"language_model.model.layers.3.self_attn.k_proj": true,
"language_model.model.layers.3.self_attn.v_proj": true,
"language_model.model.layers.3.self_attn.o_proj": true,
"language_model.model.layers.3.mlp.gate_proj": true,
"language_model.model.layers.3.mlp.down_proj": true,
"language_model.model.layers.3.mlp.up_proj": true,
"language_model.model.layers.3.input_layernorm": false,
"language_model.model.layers.3.post_attention_layernorm": false,
"language_model.model.layers.4.self_attn.q_proj": true,
"language_model.model.layers.4.self_attn.k_proj": true,
"language_model.model.layers.4.self_attn.v_proj": true,
"language_model.model.layers.4.self_attn.o_proj": true,
"language_model.model.layers.4.mlp.gate_proj": true,
"language_model.model.layers.4.mlp.down_proj": true,
"language_model.model.layers.4.mlp.up_proj": true,
"language_model.model.layers.4.input_layernorm": false,
"language_model.model.layers.4.post_attention_layernorm": false,
"language_model.model.layers.5.self_attn.q_proj": true,
"language_model.model.layers.5.self_attn.k_proj": true,
"language_model.model.layers.5.self_attn.v_proj": true,
"language_model.model.layers.5.self_attn.o_proj": true,
"language_model.model.layers.5.mlp.gate_proj": true,
"language_model.model.layers.5.mlp.down_proj": true,
"language_model.model.layers.5.mlp.up_proj": true,
"language_model.model.layers.5.input_layernorm": false,
"language_model.model.layers.5.post_attention_layernorm": false,
"language_model.model.layers.6.self_attn.q_proj": true,
"language_model.model.layers.6.self_attn.k_proj": true,
"language_model.model.layers.6.self_attn.v_proj": true,
"language_model.model.layers.6.self_attn.o_proj": true,
"language_model.model.layers.6.mlp.gate_proj": true,
"language_model.model.layers.6.mlp.down_proj": true,
"language_model.model.layers.6.mlp.up_proj": true,
"language_model.model.layers.6.input_layernorm": false,
"language_model.model.layers.6.post_attention_layernorm": false,
"language_model.model.layers.7.self_attn.q_proj": true,
"language_model.model.layers.7.self_attn.k_proj": true,
"language_model.model.layers.7.self_attn.v_proj": true,
"language_model.model.layers.7.self_attn.o_proj": true,
"language_model.model.layers.7.mlp.gate_proj": true,
"language_model.model.layers.7.mlp.down_proj": true,
"language_model.model.layers.7.mlp.up_proj": true,
"language_model.model.layers.7.input_layernorm": false,
"language_model.model.layers.7.post_attention_layernorm": false,
"language_model.model.layers.8.self_attn.q_proj": true,
"language_model.model.layers.8.self_attn.k_proj": true,
"language_model.model.layers.8.self_attn.v_proj": true,
"language_model.model.layers.8.self_attn.o_proj": true,
"language_model.model.layers.8.mlp.gate_proj": true,
"language_model.model.layers.8.mlp.down_proj": true,
"language_model.model.layers.8.mlp.up_proj": true,
"language_model.model.layers.8.input_layernorm": false,
"language_model.model.layers.8.post_attention_layernorm": false,
"language_model.model.layers.9.self_attn.q_proj": true,
"language_model.model.layers.9.self_attn.k_proj": true,
"language_model.model.layers.9.self_attn.v_proj": true,
"language_model.model.layers.9.self_attn.o_proj": true,
"language_model.model.layers.9.mlp.gate_proj": true,
"language_model.model.layers.9.mlp.down_proj": true,
"language_model.model.layers.9.mlp.up_proj": true,
"language_model.model.layers.9.input_layernorm": false,
"language_model.model.layers.9.post_attention_layernorm": false,
"language_model.model.layers.10.self_attn.q_proj": true,
"language_model.model.layers.10.self_attn.k_proj": true,
"language_model.model.layers.10.self_attn.v_proj": true,
"language_model.model.layers.10.self_attn.o_proj": true,
"language_model.model.layers.10.mlp.gate_proj": true,
"language_model.model.layers.10.mlp.down_proj": true,
"language_model.model.layers.10.mlp.up_proj": true,
"language_model.model.layers.10.input_layernorm": false,
"language_model.model.layers.10.post_attention_layernorm": false,
"language_model.model.layers.11.self_attn.q_proj": true,
"language_model.model.layers.11.self_attn.k_proj": true,
"language_model.model.layers.11.self_attn.v_proj": true,
"language_model.model.layers.11.self_attn.o_proj": true,
"language_model.model.layers.11.mlp.gate_proj": true,
"language_model.model.layers.11.mlp.down_proj": true,
"language_model.model.layers.11.mlp.up_proj": true,
"language_model.model.layers.11.input_layernorm": false,
"language_model.model.layers.11.post_attention_layernorm": false,
"language_model.model.layers.12.self_attn.q_proj": true,
"language_model.model.layers.12.self_attn.k_proj": true,
"language_model.model.layers.12.self_attn.v_proj": true,
"language_model.model.layers.12.self_attn.o_proj": true,
"language_model.model.layers.12.mlp.gate_proj": true,
"language_model.model.layers.12.mlp.down_proj": true,
"language_model.model.layers.12.mlp.up_proj": true,
"language_model.model.layers.12.input_layernorm": false,
"language_model.model.layers.12.post_attention_layernorm": false,
"language_model.model.layers.13.self_attn.q_proj": true,
"language_model.model.layers.13.self_attn.k_proj": true,
"language_model.model.layers.13.self_attn.v_proj": true,
"language_model.model.layers.13.self_attn.o_proj": true,
"language_model.model.layers.13.mlp.gate_proj": true,
"language_model.model.layers.13.mlp.down_proj": true,
"language_model.model.layers.13.mlp.up_proj": true,
"language_model.model.layers.13.input_layernorm": false,
"language_model.model.layers.13.post_attention_layernorm": false,
"language_model.model.layers.14.self_attn.q_proj": true,
"language_model.model.layers.14.self_attn.k_proj": true,
"language_model.model.layers.14.self_attn.v_proj": true,
"language_model.model.layers.14.self_attn.o_proj": true,
"language_model.model.layers.14.mlp.gate_proj": true,
"language_model.model.layers.14.mlp.down_proj": true,
"language_model.model.layers.14.mlp.up_proj": true,
"language_model.model.layers.14.input_layernorm": false,
"language_model.model.layers.14.post_attention_layernorm": false,
"language_model.model.layers.15.self_attn.q_proj": true,
"language_model.model.layers.15.self_attn.k_proj": true,
"language_model.model.layers.15.self_attn.v_proj": true,
"language_model.model.layers.15.self_attn.o_proj": true,
"language_model.model.layers.15.mlp.gate_proj": true,
"language_model.model.layers.15.mlp.down_proj": true,
"language_model.model.layers.15.mlp.up_proj": true,
"language_model.model.layers.15.input_layernorm": false,
"language_model.model.layers.15.post_attention_layernorm": false,
"language_model.model.layers.16.self_attn.q_proj": true,
"language_model.model.layers.16.self_attn.k_proj": true,
"language_model.model.layers.16.self_attn.v_proj": true,
"language_model.model.layers.16.self_attn.o_proj": true,
"language_model.model.layers.16.mlp.gate_proj": true,
"language_model.model.layers.16.mlp.down_proj": true,
"language_model.model.layers.16.mlp.up_proj": true,
"language_model.model.layers.16.input_layernorm": false,
"language_model.model.layers.16.post_attention_layernorm": false,
"language_model.model.layers.17.self_attn.q_proj": true,
"language_model.model.layers.17.self_attn.k_proj": true,
"language_model.model.layers.17.self_attn.v_proj": true,
"language_model.model.layers.17.self_attn.o_proj": true,
"language_model.model.layers.17.mlp.gate_proj": true,
"language_model.model.layers.17.mlp.down_proj": true,
"language_model.model.layers.17.mlp.up_proj": true,
"language_model.model.layers.17.input_layernorm": false,
"language_model.model.layers.17.post_attention_layernorm": false,
"language_model.model.layers.18.self_attn.q_proj": true,
"language_model.model.layers.18.self_attn.k_proj": true,
"language_model.model.layers.18.self_attn.v_proj": true,
"language_model.model.layers.18.self_attn.o_proj": true,
"language_model.model.layers.18.mlp.gate_proj": true,
"language_model.model.layers.18.mlp.down_proj": true,
"language_model.model.layers.18.mlp.up_proj": true,
"language_model.model.layers.18.input_layernorm": false,
"language_model.model.layers.18.post_attention_layernorm": false,
"language_model.model.layers.19.self_attn.q_proj": true,
"language_model.model.layers.19.self_attn.k_proj": true,
"language_model.model.layers.19.self_attn.v_proj": true,
"language_model.model.layers.19.self_attn.o_proj": true,
"language_model.model.layers.19.mlp.gate_proj": true,
"language_model.model.layers.19.mlp.down_proj": true,
"language_model.model.layers.19.mlp.up_proj": true,
"language_model.model.layers.19.input_layernorm": false,
"language_model.model.layers.19.post_attention_layernorm": false,
"language_model.model.layers.20.self_attn.q_proj": true,
"language_model.model.layers.20.self_attn.k_proj": true,
"language_model.model.layers.20.self_attn.v_proj": true,
"language_model.model.layers.20.self_attn.o_proj": true,
"language_model.model.layers.20.mlp.gate_proj": true,
"language_model.model.layers.20.mlp.down_proj": true,
"language_model.model.layers.20.mlp.up_proj": true,
"language_model.model.layers.20.input_layernorm": false,
"language_model.model.layers.20.post_attention_layernorm": false,
"language_model.model.layers.21.self_attn.q_proj": true,
"language_model.model.layers.21.self_attn.k_proj": true,
"language_model.model.layers.21.self_attn.v_proj": true,
"language_model.model.layers.21.self_attn.o_proj": true,
"language_model.model.layers.21.mlp.gate_proj": true,
"language_model.model.layers.21.mlp.down_proj": true,
"language_model.model.layers.21.mlp.up_proj": true,
"language_model.model.layers.21.input_layernorm": false,
"language_model.model.layers.21.post_attention_layernorm": false,
"language_model.model.layers.22.self_attn.q_proj": true,
"language_model.model.layers.22.self_attn.k_proj": true,
"language_model.model.layers.22.self_attn.v_proj": true,
"language_model.model.layers.22.self_attn.o_proj": true,
"language_model.model.layers.22.mlp.gate_proj": true,
"language_model.model.layers.22.mlp.down_proj": true,
"language_model.model.layers.22.mlp.up_proj": true,
"language_model.model.layers.22.input_layernorm": false,
"language_model.model.layers.22.post_attention_layernorm": false,
"language_model.model.layers.23.self_attn.q_proj": true,
"language_model.model.layers.23.self_attn.k_proj": true,
"language_model.model.layers.23.self_attn.v_proj": true,
"language_model.model.layers.23.self_attn.o_proj": true,
"language_model.model.layers.23.mlp.gate_proj": true,
"language_model.model.layers.23.mlp.down_proj": true,
"language_model.model.layers.23.mlp.up_proj": true,
"language_model.model.layers.23.input_layernorm": false,
"language_model.model.layers.23.post_attention_layernorm": false,
"language_model.model.layers.24.self_attn.q_proj": true,
"language_model.model.layers.24.self_attn.k_proj": true,
"language_model.model.layers.24.self_attn.v_proj": true,
"language_model.model.layers.24.self_attn.o_proj": true,
"language_model.model.layers.24.mlp.gate_proj": true,
"language_model.model.layers.24.mlp.down_proj": true,
"language_model.model.layers.24.mlp.up_proj": true,
"language_model.model.layers.24.input_layernorm": false,
"language_model.model.layers.24.post_attention_layernorm": false,
"language_model.model.layers.25.self_attn.q_proj": true,
"language_model.model.layers.25.self_attn.k_proj": true,
"language_model.model.layers.25.self_attn.v_proj": true,
"language_model.model.layers.25.self_attn.o_proj": true,
"language_model.model.layers.25.mlp.gate_proj": true,
"language_model.model.layers.25.mlp.down_proj": true,
"language_model.model.layers.25.mlp.up_proj": true,
"language_model.model.layers.25.input_layernorm": false,
"language_model.model.layers.25.post_attention_layernorm": false,
"language_model.model.layers.26.self_attn.q_proj": true,
"language_model.model.layers.26.self_attn.k_proj": true,
"language_model.model.layers.26.self_attn.v_proj": true,
"language_model.model.layers.26.self_attn.o_proj": true,
"language_model.model.layers.26.mlp.gate_proj": true,
"language_model.model.layers.26.mlp.down_proj": true,
"language_model.model.layers.26.mlp.up_proj": true,
"language_model.model.layers.26.input_layernorm": false,
"language_model.model.layers.26.post_attention_layernorm": false,
"language_model.model.layers.27.self_attn.q_proj": true,
"language_model.model.layers.27.self_attn.k_proj": true,
"language_model.model.layers.27.self_attn.v_proj": true,
"language_model.model.layers.27.self_attn.o_proj": true,
"language_model.model.layers.27.mlp.gate_proj": true,
"language_model.model.layers.27.mlp.down_proj": true,
"language_model.model.layers.27.mlp.up_proj": true,
"language_model.model.layers.27.input_layernorm": false,
"language_model.model.layers.27.post_attention_layernorm": false,
"language_model.model.norm": false,
"language_model.lm_head": true
},
"remove_invalid_values": false,
"repetition_penalty": 1.0,
"return_dict": true,
"return_dict_in_generate": false,
"rms_norm_eps": 1e-06,
"rope_scaling": {
"mrope_section": [
16,
24,
24
],
"rope_type": "default",
"type": "default"
},
"rope_theta": 1000000.0,
"sep_token_id": null,
"sliding_window": 32768,
"suppress_tokens": null,
"task_specific_params": null,
"temperature": 1.0,
"text_config": {
"vocab_size": 152064,
"max_position_embeddings": 128000,
"hidden_size": 3584,
"intermediate_size": 18944,
"num_hidden_layers": 28,
"num_attention_heads": 28,
"use_sliding_window": false,
"sliding_window": null,
"max_window_layers": 28,
"num_key_value_heads": 4,
"hidden_act": "silu",
"initializer_range": 0.02,
"rms_norm_eps": 1e-06,
"use_cache": false,
"rope_theta": 1000000.0,
"attention_dropout": 0.0,
"rope_scaling": {
"mrope_section": [
16,
24,
24
],
"rope_type": "default",
"type": "default"
},
"layer_types": [
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention",
"full_attention"
],
"image_token_id": null,
"video_token_id": null,
"return_dict": true,
"output_hidden_states": false,
"torchscript": false,
"torch_dtype": "bfloat16",
"pruned_heads": {},
"tie_word_embeddings": false,
"chunk_size_feed_forward": 0,
"is_encoder_decoder": false,
"is_decoder": false,
"cross_attention_hidden_size": null,
"add_cross_attention": false,
"tie_encoder_decoder": false,
"architectures": [
"Qwen2_5_VLForConditionalGeneration"
],
"finetuning_task": null,
"id2label": {
"0": "LABEL_0",
"1": "LABEL_1"
},
"label2id": {
"LABEL_0": 0,
"LABEL_1": 1
},
"task_specific_params": null,
"problem_type": null,
"tokenizer_class": null,
"prefix": null,
"bos_token_id": 151643,
"pad_token_id": null,
"eos_token_id": 151645,
"sep_token_id": null,
"decoder_start_token_id": null,
"max_length": 20,
"min_length": 0,
"do_sample": false,
"early_stopping": false,
"num_beams": 1,
"num_beam_groups": 1,
"diversity_penalty": 0.0,
"temperature": 1.0,
"top_k": 50,
"top_p": 1.0,
"typical_p": 1.0,
"repetition_penalty": 1.0,
"length_penalty": 1.0,
"no_repeat_ngram_size": 0,
"encoder_no_repeat_ngram_size": 0,
"bad_words_ids": null,
"num_return_sequences": 1,
"output_scores": false,
"return_dict_in_generate": false,
"forced_bos_token_id": null,
"forced_eos_token_id": null,
"remove_invalid_values": false,
"exponential_decay_length_penalty": null,
"suppress_tokens": null,
"begin_suppress_tokens": null,
"_name_or_path": "/home/notebook/data/group/wth/QualityReasoning/Models/Qwen2.5-VL-7B-Instruct",
"model_type": "qwen2_5_vl_text",
"vision_end_token_id": 151653,
"vision_start_token_id": 151652,
"vision_token_id": 151654,
"tf_legacy_loss": false,
"use_bfloat16": false,
"output_attentions": false
},
"tf_legacy_loss": false,
"tie_encoder_decoder": false,
"tie_word_embeddings": false,
"tokenizer_class": null,
"top_k": 50,
"top_p": 1.0,
"torchscript": false,
"transformers_version": "4.54.1",
"typical_p": 1.0,
"use_bfloat16": false,
"use_cache": false,
"use_sliding_window": false,
"video_token_id": 151656,
"vision_config": {
"return_dict": true,
"output_hidden_states": false,
"torchscript": false,
"torch_dtype": "bfloat16",
"pruned_heads": {},
"tie_word_embeddings": true,
"chunk_size_feed_forward": 0,
"is_encoder_decoder": false,
"is_decoder": false,
"cross_attention_hidden_size": null,
"add_cross_attention": false,
"tie_encoder_decoder": false,
"architectures": null,
"finetuning_task": null,
"id2label": {
"0": "LABEL_0",
"1": "LABEL_1"
},
"label2id": {
"LABEL_0": 0,
"LABEL_1": 1
},
"task_specific_params": null,
"problem_type": null,
"tokenizer_class": null,
"prefix": null,
"bos_token_id": null,
"pad_token_id": null,
"eos_token_id": null,
"sep_token_id": null,
"decoder_start_token_id": null,
"max_length": 20,
"min_length": 0,
"do_sample": false,
"early_stopping": false,
"num_beams": 1,
"num_beam_groups": 1,
"diversity_penalty": 0.0,
"temperature": 1.0,
"top_k": 50,
"top_p": 1.0,
"typical_p": 1.0,
"repetition_penalty": 1.0,
"length_penalty": 1.0,
"no_repeat_ngram_size": 0,
"encoder_no_repeat_ngram_size": 0,
"bad_words_ids": null,
"num_return_sequences": 1,
"output_scores": false,
"return_dict_in_generate": false,
"forced_bos_token_id": null,
"forced_eos_token_id": null,
"remove_invalid_values": false,
"exponential_decay_length_penalty": null,
"suppress_tokens": null,
"begin_suppress_tokens": null,
"_name_or_path": "",
"in_chans": 3,
"model_type": "qwen2_5_vl",
"spatial_patch_size": 14,
"tf_legacy_loss": false,
"use_bfloat16": false,
"depth": 32,
"hidden_size": 1280,
"hidden_act": "silu",
"intermediate_size": 3420,
"num_heads": 16,
"in_channels": 3,
"patch_size": 14,
"spatial_merge_size": 2,
"temporal_patch_size": 2,
"tokens_per_second": 2,
"window_size": 112,
"fullatt_block_indexes": [
7,
15,
23,
31
],
"out_hidden_size": 3584,
"initializer_range": 0.02,
"output_attentions": false
},
"vision_end_token_id": 151653,
"vision_start_token_id": 151652,
"vision_token_id": 151654,
"vocab_size": 152064
}