IntMeGroup/LMM4LMM_8B_MOSs2_st2

1{2  "_commit_hash": null,3  "_name_or_path": "weights/stage1/stage1s_mos2",4  "architectures": [5    "InternVLChatModel"6  ],7  "auto_map": {8    "AutoConfig": "configuration_internvl_chat.InternVLChatConfig",9    "AutoModel": "modeling_internvl_chat.InternVLChatModel",10    "AutoModelForCausalLM": "modeling_internvl_chat.InternVLChatModel"11  },12  "downsample_ratio": 0.5,13  "dynamic_image_size": true,14  "force_image_size": 448,15  "llm_config": {16    "_name_or_path": "internlm/internlm2_5-7b-chat",17    "add_cross_attention": false,18    "architectures": [19      "InternLM2ForCausalLM"20    ],21    "attn_implementation": "flash_attention_2",22    "auto_map": {23      "AutoConfig": "configuration_internlm2.InternLM2Config",24      "AutoModel": "modeling_internlm2.InternLM2ForCausalLM",25      "AutoModelForCausalLM": "modeling_internlm2.InternLM2ForCausalLM",26      "AutoModelForSequenceClassification": "modeling_internlm2.InternLM2ForSequenceClassification"27    },28    "bad_words_ids": null,29    "begin_suppress_tokens": null,30    "bias": false,31    "bos_token_id": 1,32    "chunk_size_feed_forward": 0,33    "cross_attention_hidden_size": null,34    "decoder_start_token_id": null,35    "diversity_penalty": 0.0,36    "do_sample": false,37    "early_stopping": false,38    "encoder_no_repeat_ngram_size": 0,39    "eos_token_id": 2,40    "exponential_decay_length_penalty": null,41    "finetuning_task": null,42    "forced_bos_token_id": null,43    "forced_eos_token_id": null,44    "hidden_act": "silu",45    "hidden_size": 4096,46    "id2label": {47      "0": "LABEL_0",48      "1": "LABEL_1"49    },50    "initializer_range": 0.02,51    "intermediate_size": 14336,52    "is_decoder": false,53    "is_encoder_decoder": false,54    "label2id": {55      "LABEL_0": 0,56      "LABEL_1": 157    },58    "length_penalty": 1.0,59    "max_length": 20,60    "max_position_embeddings": 32768,61    "min_length": 0,62    "model_type": "internlm2",63    "no_repeat_ngram_size": 0,64    "num_attention_heads": 32,65    "num_beam_groups": 1,66    "num_beams": 1,67    "num_hidden_layers": 32,68    "num_key_value_heads": 8,69    "num_return_sequences": 1,70    "output_attentions": false,71    "output_hidden_states": false,72    "output_scores": false,73    "pad_token_id": 2,74    "prefix": null,75    "pretraining_tp": 1,76    "problem_type": null,77    "pruned_heads": {},78    "remove_invalid_values": false,79    "repetition_penalty": 1.0,80    "return_dict": true,81    "return_dict_in_generate": false,82    "rms_norm_eps": 1e-05,83    "rope_scaling": {84      "factor": 2.0,85      "type": "dynamic"86    },87    "rope_theta": 1000000,88    "sep_token_id": null,89    "suppress_tokens": null,90    "task_specific_params": null,91    "temperature": 1.0,92    "tf_legacy_loss": false,93    "tie_encoder_decoder": false,94    "tie_word_embeddings": false,95    "tokenizer_class": null,96    "top_k": 50,97    "top_p": 1.0,98    "torch_dtype": "bfloat16",99    "torchscript": false,100    "transformers_version": "4.37.2",101    "typical_p": 1.0,102    "use_bfloat16": true,103    "use_cache": false,104    "vocab_size": 92553105  },106  "max_dynamic_patch": 6,107  "min_dynamic_patch": 1,108  "model_type": "internvl_chat",109  "pad2square": false,110  "ps_version": "v2",111  "select_layer": -1,112  "template": "internlm2-chat",113  "torch_dtype": "bfloat16",114  "transformers_version": null,115  "use_backbone_lora": 8,116  "use_llm_lora": 8,117  "use_thumbnail": true,118  "vision_config": {119    "_name_or_path": "",120    "add_cross_attention": false,121    "architectures": [122      "InternVisionModel"123    ],124    "attention_dropout": 0.0,125    "bad_words_ids": null,126    "begin_suppress_tokens": null,127    "bos_token_id": null,128    "chunk_size_feed_forward": 0,129    "cross_attention_hidden_size": null,130    "decoder_start_token_id": null,131    "diversity_penalty": 0.0,132    "do_sample": false,133    "drop_path_rate": 0.1,134    "dropout": 0.0,135    "early_stopping": false,136    "encoder_no_repeat_ngram_size": 0,137    "eos_token_id": null,138    "exponential_decay_length_penalty": null,139    "finetuning_task": null,140    "forced_bos_token_id": null,141    "forced_eos_token_id": null,142    "hidden_act": "gelu",143    "hidden_size": 1024,144    "id2label": {145      "0": "LABEL_0",146      "1": "LABEL_1"147    },148    "image_size": 448,149    "initializer_factor": 1.0,150    "initializer_range": 0.02,151    "intermediate_size": 4096,152    "is_decoder": false,153    "is_encoder_decoder": false,154    "label2id": {155      "LABEL_0": 0,156      "LABEL_1": 1157    },158    "layer_norm_eps": 1e-06,159    "length_penalty": 1.0,160    "max_length": 20,161    "min_length": 0,162    "model_type": "intern_vit_6b",163    "no_repeat_ngram_size": 0,164    "norm_type": "layer_norm",165    "num_attention_heads": 16,166    "num_beam_groups": 1,167    "num_beams": 1,168    "num_channels": 3,169    "num_hidden_layers": 24,170    "num_return_sequences": 1,171    "output_attentions": false,172    "output_hidden_states": false,173    "output_scores": false,174    "pad_token_id": null,175    "patch_size": 14,176    "prefix": null,177    "problem_type": null,178    "pruned_heads": {},179    "qk_normalization": false,180    "qkv_bias": true,181    "remove_invalid_values": false,182    "repetition_penalty": 1.0,183    "return_dict": true,184    "return_dict_in_generate": false,185    "sep_token_id": null,186    "suppress_tokens": null,187    "task_specific_params": null,188    "temperature": 1.0,189    "tf_legacy_loss": false,190    "tie_encoder_decoder": false,191    "tie_word_embeddings": true,192    "tokenizer_class": null,193    "top_k": 50,194    "top_p": 1.0,195    "torch_dtype": "bfloat16",196    "torchscript": false,197    "transformers_version": "4.37.2",198    "typical_p": 1.0,199    "use_bfloat16": true,200    "use_flash_attn": true201  }202}203