| { | |
| "_name_or_path": "/tmp/tmp_w31aofq", | |
| "architectures": [ | |
| "LlamaForCausalLM" | |
| ], | |
| "attention_bias": false, | |
| "attention_dropout": 0.0, | |
| "bos_token_id": 0, | |
| "eos_token_id": 2, | |
| "hidden_act": "silu", | |
| "hidden_size": 768, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 3072, | |
| "max_position_embeddings": 2048, | |
| "model_type": "llama", | |
| "num_attention_heads": 12, | |
| "num_hidden_layers": 12, | |
| "num_key_value_heads": 12, | |
| "pad_token_id": 1, | |
| "pretraining_tp": 1, | |
| "quant_config": { | |
| "default": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "model_layer_0": { | |
| "mlp": { | |
| "down_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "gate_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "up_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| } | |
| }, | |
| "self_attn": { | |
| "k_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "matmul_0": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "matmul_1": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "o_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "q_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "v_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| } | |
| } | |
| }, | |
| "model_layer_1": { | |
| "mlp": { | |
| "down_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "gate_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "up_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| } | |
| }, | |
| "self_attn": { | |
| "k_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "matmul_0": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "matmul_1": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "o_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "q_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "v_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| } | |
| } | |
| }, | |
| "model_layer_10": { | |
| "mlp": { | |
| "down_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "gate_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "up_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| } | |
| }, | |
| "self_attn": { | |
| "k_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "matmul_0": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "matmul_1": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "o_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "q_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "v_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| } | |
| } | |
| }, | |
| "model_layer_11": { | |
| "mlp": { | |
| "down_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "gate_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "up_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| } | |
| }, | |
| "self_attn": { | |
| "k_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "matmul_0": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "matmul_1": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "o_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "q_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "v_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| } | |
| } | |
| }, | |
| "model_layer_2": { | |
| "mlp": { | |
| "down_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "gate_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "up_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| } | |
| }, | |
| "self_attn": { | |
| "k_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "matmul_0": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "matmul_1": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "o_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "q_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "v_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| } | |
| } | |
| }, | |
| "model_layer_3": { | |
| "mlp": { | |
| "down_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "gate_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "up_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| } | |
| }, | |
| "self_attn": { | |
| "k_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "matmul_0": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "matmul_1": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "o_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "q_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "v_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| } | |
| } | |
| }, | |
| "model_layer_4": { | |
| "mlp": { | |
| "down_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "gate_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "up_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| } | |
| }, | |
| "self_attn": { | |
| "k_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "matmul_0": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "matmul_1": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "o_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "q_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "v_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| } | |
| } | |
| }, | |
| "model_layer_5": { | |
| "mlp": { | |
| "down_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "gate_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "up_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| } | |
| }, | |
| "self_attn": { | |
| "k_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "matmul_0": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "matmul_1": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "o_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "q_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "v_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| } | |
| } | |
| }, | |
| "model_layer_6": { | |
| "mlp": { | |
| "down_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "gate_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "up_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| } | |
| }, | |
| "self_attn": { | |
| "k_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "matmul_0": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "matmul_1": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "o_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "q_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "v_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| } | |
| } | |
| }, | |
| "model_layer_7": { | |
| "mlp": { | |
| "down_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "gate_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "up_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| } | |
| }, | |
| "self_attn": { | |
| "k_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "matmul_0": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "matmul_1": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "o_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "q_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "v_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| } | |
| } | |
| }, | |
| "model_layer_8": { | |
| "mlp": { | |
| "down_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "gate_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "up_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| } | |
| }, | |
| "self_attn": { | |
| "k_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "matmul_0": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "matmul_1": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "o_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "q_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "v_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| } | |
| } | |
| }, | |
| "model_layer_9": { | |
| "mlp": { | |
| "down_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "gate_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "up_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| } | |
| }, | |
| "self_attn": { | |
| "k_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "matmul_0": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "matmul_1": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "o_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "q_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| }, | |
| "v_proj": { | |
| "bypass": true, | |
| "name": "integer" | |
| } | |
| } | |
| } | |
| }, | |
| "quantization_config": { | |
| "bnb_4bit_compute_dtype": "bfloat16", | |
| "bnb_4bit_quant_type": "fp4", | |
| "bnb_4bit_use_double_quant": true, | |
| "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" | |
| }, | |
| "rms_norm_eps": 1e-06, | |
| "rope_scaling": null, | |
| "rope_theta": 10000.0, | |
| "tie_word_embeddings": false, | |
| "torch_dtype": "float16", | |
| "transformers_version": "4.37.1", | |
| "use_cache": true, | |
| "vocab_size": 32000 | |
| } | |