Delete config.json
Browse files- config.json +0 -110
config.json
DELETED
|
@@ -1,110 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"architectures": [
|
| 3 |
-
"Qwen3ForCausalLM"
|
| 4 |
-
],
|
| 5 |
-
"attention_bias": false,
|
| 6 |
-
"attention_dropout": 0.0,
|
| 7 |
-
"bos_token_id": 151643,
|
| 8 |
-
"eos_token_id": 151645,
|
| 9 |
-
"head_dim": 128,
|
| 10 |
-
"hidden_act": "silu",
|
| 11 |
-
"hidden_size": 2560,
|
| 12 |
-
"initializer_range": 0.02,
|
| 13 |
-
"intermediate_size": 9728,
|
| 14 |
-
"max_position_embeddings": 40960,
|
| 15 |
-
"max_window_layers": 36,
|
| 16 |
-
"model_type": "qwen3",
|
| 17 |
-
"num_attention_heads": 32,
|
| 18 |
-
"num_hidden_layers": 36,
|
| 19 |
-
"num_key_value_heads": 8,
|
| 20 |
-
"quantization_config": {
|
| 21 |
-
"include_input_output_embeddings": false,
|
| 22 |
-
"modules_to_not_convert": [],
|
| 23 |
-
"quant_method": "torchao",
|
| 24 |
-
"quant_type": {
|
| 25 |
-
"default": {
|
| 26 |
-
"_data": {
|
| 27 |
-
"module_fqn_to_config": {
|
| 28 |
-
"_default": {
|
| 29 |
-
"_data": {
|
| 30 |
-
"act_mapping_type": {
|
| 31 |
-
"_data": "ASYMMETRIC",
|
| 32 |
-
"_type": "MappingType"
|
| 33 |
-
},
|
| 34 |
-
"layout": {
|
| 35 |
-
"_data": {},
|
| 36 |
-
"_type": "QDQLayout",
|
| 37 |
-
"_version": 1
|
| 38 |
-
},
|
| 39 |
-
"weight_dtype": {
|
| 40 |
-
"_data": "int4",
|
| 41 |
-
"_type": "torch.dtype"
|
| 42 |
-
},
|
| 43 |
-
"weight_granularity": {
|
| 44 |
-
"_data": {
|
| 45 |
-
"group_size": 32
|
| 46 |
-
},
|
| 47 |
-
"_type": "PerGroup",
|
| 48 |
-
"_version": 1
|
| 49 |
-
},
|
| 50 |
-
"weight_mapping_type": {
|
| 51 |
-
"_data": "SYMMETRIC",
|
| 52 |
-
"_type": "MappingType"
|
| 53 |
-
},
|
| 54 |
-
"weight_scale_dtype": {
|
| 55 |
-
"_data": "bfloat16",
|
| 56 |
-
"_type": "torch.dtype"
|
| 57 |
-
}
|
| 58 |
-
},
|
| 59 |
-
"_type": "Int8DynamicActivationIntxWeightConfig",
|
| 60 |
-
"_version": 1
|
| 61 |
-
},
|
| 62 |
-
"model.embed_tokens": {
|
| 63 |
-
"_data": {
|
| 64 |
-
"granularity": {
|
| 65 |
-
"_data": {
|
| 66 |
-
"axis": 0
|
| 67 |
-
},
|
| 68 |
-
"_type": "PerAxis",
|
| 69 |
-
"_version": 1
|
| 70 |
-
},
|
| 71 |
-
"layout": {
|
| 72 |
-
"_data": {},
|
| 73 |
-
"_type": "QDQLayout",
|
| 74 |
-
"_version": 1
|
| 75 |
-
},
|
| 76 |
-
"mapping_type": {
|
| 77 |
-
"_data": "SYMMETRIC",
|
| 78 |
-
"_type": "MappingType"
|
| 79 |
-
},
|
| 80 |
-
"scale_dtype": null,
|
| 81 |
-
"weight_dtype": {
|
| 82 |
-
"_data": "int8",
|
| 83 |
-
"_type": "torch.dtype"
|
| 84 |
-
}
|
| 85 |
-
},
|
| 86 |
-
"_type": "IntxWeightOnlyConfig",
|
| 87 |
-
"_version": 1
|
| 88 |
-
}
|
| 89 |
-
}
|
| 90 |
-
},
|
| 91 |
-
"_type": "ModuleFqnToConfig",
|
| 92 |
-
"_version": 1
|
| 93 |
-
}
|
| 94 |
-
},
|
| 95 |
-
"quant_type_kwargs": {
|
| 96 |
-
"include_embedding": true
|
| 97 |
-
},
|
| 98 |
-
"untie_embedding_weights": true
|
| 99 |
-
},
|
| 100 |
-
"rms_norm_eps": 1e-06,
|
| 101 |
-
"rope_scaling": null,
|
| 102 |
-
"rope_theta": 1000000,
|
| 103 |
-
"sliding_window": null,
|
| 104 |
-
"tie_word_embeddings": false,
|
| 105 |
-
"torch_dtype": "float32",
|
| 106 |
-
"transformers_version": "4.52.3",
|
| 107 |
-
"use_cache": true,
|
| 108 |
-
"use_sliding_window": false,
|
| 109 |
-
"vocab_size": 151936
|
| 110 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|