| { | |
| "_name_or_path": "/ceph/hdd/staff/charpent/.cache/modelsfsrcdvi92q8mrsw8", | |
| "architectures": [ | |
| "GitForCausalLM" | |
| ], | |
| "attention_probs_dropout_prob": 0.1, | |
| "bos_token_id": 101, | |
| "classifier_dropout": null, | |
| "eos_token_id": 102, | |
| "hidden_act": "gelu", | |
| "hidden_dropout_prob": 0.1, | |
| "hidden_size": 768, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 3072, | |
| "layer_norm_eps": 1e-12, | |
| "max_position_embeddings": 1024, | |
| "model_type": "git", | |
| "num_attention_heads": 12, | |
| "num_hidden_layers": 6, | |
| "num_image_with_embedding": null, | |
| "pad_token_id": 0, | |
| "position_embedding_type": "absolute", | |
| "quantization_config": { | |
| "_load_in_4bit": true, | |
| "_load_in_8bit": false, | |
| "bnb_4bit_compute_dtype": "bfloat16", | |
| "bnb_4bit_quant_storage": "uint8", | |
| "bnb_4bit_quant_type": "fp4", | |
| "bnb_4bit_use_double_quant": false, | |
| "llm_int8_enable_fp32_cpu_offload": false, | |
| "llm_int8_has_fp16_weight": false, | |
| "llm_int8_skip_modules": [ | |
| "lm_head" | |
| ], | |
| "llm_int8_threshold": 6.0, | |
| "load_in_4bit": true, | |
| "load_in_8bit": false, | |
| "quant_method": "bitsandbytes" | |
| }, | |
| "tie_word_embeddings": false, | |
| "torch_dtype": "float16", | |
| "transformers_version": "4.40.0", | |
| "use_cache": true, | |
| "vision_config": { | |
| "dropout": 0.0, | |
| "image_size": 480, | |
| "initializer_factor": 1.0, | |
| "model_type": "git_vision_model", | |
| "projection_dim": 512 | |
| }, | |
| "vocab_size": 30522 | |
| } | |