Upload config.json

config.json

@@ -0,0 +1,46 @@
+{
+ "_name_or_path": "model/videollama2_7b_spatial_slot_init/",
+ "architectures": [
+ "Videollama2MistralForCausalLM"
+ ],
+ "attention_dropout": 0.0,
+ "bos_token_id": 1,
+ "downsample_num": 1,
+ "eos_token_id": 2,
+ "freeze_mm_mlp_adapter": false,
+ "hidden_act": "silu",
+ "hidden_size": 4096,
+ "image_aspect_ratio": "pad",
+ "initializer_range": 0.02,
+ "intermediate_size": 14336,
+ "max_position_embeddings": 32768,
+ "mm_hidden_size": 1024,
+ "mm_projector_lr": null,
+ "mm_projector_type": "spatial_slot",
+ "mm_use_im_patch_token": false,
+ "mm_use_im_start_end": false,
+ "mm_vision_select_feature": "patch",
+ "mm_vision_select_layer": -2,
+ "mm_vision_tower": "model/clip-vit-large-patch14-336",
+ "model_type": "trace_mistral",
+ "num_attention_heads": 32,
+ "num_frames": 128,
+ "num_hidden_layers": 32,
+ "num_key_value_heads": 8,
+ "rms_norm_eps": 1e-05,
+ "rope_theta": 1000000.0,
+ "score_vocab_size": 13,
+ "sliding_window": null,
+ "tie_word_embeddings": false,
+ "time_vocab_size": 13,
+ "tokenizer_model_max_length": 4096,
+ "tokenizer_padding_side": "right",
+ "torch_dtype": "bfloat16",
+ "transformers_version": "4.36.2",
+ "tune_lm_embed_head": false,
+ "tune_mm_embed_head": true,
+ "tune_mm_mlp_adapter": true,
+ "use_cache": true,
+ "use_mm_proj": true,
+ "vocab_size": 32000
+}