tbhrc/phi_3_5_mini_instruct_4bit

1{2    "architectures": [3        "Phi3ForCausalLM"4    ],5    "attention_bias": false,6    "attention_dropout": 0.0,7    "auto_map": {8        "AutoConfig": "configuration_phi3.Phi3Config",9        "AutoModelForCausalLM": "modeling_phi3.Phi3ForCausalLM"10    },11    "bos_token_id": 1,12    "embd_pdrop": 0.0,13    "eos_token_id": 32000,14    "hidden_act": "silu",15    "hidden_size": 3072,16    "initializer_range": 0.02,17    "intermediate_size": 8192,18    "max_position_embeddings": 131072,19    "model_type": "phi3",20    "num_attention_heads": 32,21    "num_hidden_layers": 32,22    "num_key_value_heads": 32,23    "original_max_position_embeddings": 4096,24    "pad_token_id": 32000,25    "quantization": {26        "group_size": 64,27        "bits": 428    },29    "resid_pdrop": 0.0,30    "rms_norm_eps": 1e-05,31    "rope_scaling": {32        "long_factor": [33            1.0800000429153442,34            1.1100000143051147,35            1.1399999856948853,36            1.340000033378601,37            1.5899999141693115,38            1.600000023841858,39            1.6200000047683716,40            2.620000123977661,41            3.2300000190734863,42            3.2300000190734863,43            4.789999961853027,44            7.400000095367432,45            7.700000286102295,46            9.09000015258789,47            12.199999809265137,48            17.670000076293945,49            24.46000099182129,50            28.57000160217285,51            30.420001983642578,52            30.840002059936523,53            32.590003967285156,54            32.93000411987305,55            42.320003509521484,56            44.96000289916992,57            50.340003967285156,58            50.45000457763672,59            57.55000305175781,60            57.93000411987305,61            58.21000289916992,62            60.1400032043457,63            62.61000442504883,64            62.62000274658203,65            62.71000289916992,66            63.1400032043457,67            63.1400032043457,68            63.77000427246094,69            63.93000411987305,70            63.96000289916992,71            63.970001220703125,72            64.02999877929688,73            64.06999969482422,74            64.08000183105469,75            64.12000274658203,76            64.41000366210938,77            64.4800033569336,78            64.51000213623047,79            64.52999877929688,80            64.8399963378906281        ],82        "short_factor": [83            1.0,84            1.0199999809265137,85            1.0299999713897705,86            1.0299999713897705,87            1.0499999523162842,88            1.0499999523162842,89            1.0499999523162842,90            1.0499999523162842,91            1.0499999523162842,92            1.0699999332427979,93            1.0999999046325684,94            1.1099998950958252,95            1.1599998474121094,96            1.1599998474121094,97            1.1699998378753662,98            1.2899998426437378,99            1.339999794960022,100            1.679999828338623,101            1.7899998426437378,102            1.8199998140335083,103            1.8499997854232788,104            1.8799997568130493,105            1.9099997282028198,106            1.9399996995925903,107            1.9899996519088745,108            2.0199997425079346,109            2.0199997425079346,110            2.0199997425079346,111            2.0199997425079346,112            2.0199997425079346,113            2.0199997425079346,114            2.0299997329711914,115            2.0299997329711914,116            2.0299997329711914,117            2.0299997329711914,118            2.0299997329711914,119            2.0299997329711914,120            2.0299997329711914,121            2.0299997329711914,122            2.0299997329711914,123            2.0799996852874756,124            2.0899996757507324,125            2.189999580383301,126            2.2199995517730713,127            2.5899994373321533,128            2.729999542236328,129            2.749999523162842,130            2.8399994373321533131        ],132        "type": "longrope"133    },134    "rope_theta": 10000.0,135    "sliding_window": 262144,136    "tie_word_embeddings": false,137    "torch_dtype": "bfloat16",138    "transformers_version": "4.43.3",139    "use_cache": true,140    "vocab_size": 32064141}