Upload quantized ONNX model - config.json
Browse files- config.json +81 -0
config.json
ADDED
|
@@ -0,0 +1,81 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "nvidia/prompt-task-and-complexity-classifier",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"DebertaV2Model"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"hidden_act": "gelu",
|
| 8 |
+
"hidden_dropout_prob": 0.1,
|
| 9 |
+
"hidden_size": 768,
|
| 10 |
+
"initializer_range": 0.02,
|
| 11 |
+
"intermediate_size": 3072,
|
| 12 |
+
"layer_norm_eps": 1e-07,
|
| 13 |
+
"max_position_embeddings": 512,
|
| 14 |
+
"model_type": "deberta-v2",
|
| 15 |
+
"num_attention_heads": 12,
|
| 16 |
+
"num_hidden_layers": 12,
|
| 17 |
+
"pad_token_id": 0,
|
| 18 |
+
"pooler_dropout": 0.0,
|
| 19 |
+
"pooler_hidden_act": "gelu",
|
| 20 |
+
"pooler_hidden_size": 768,
|
| 21 |
+
"position_biased_input": false,
|
| 22 |
+
"position_buckets": 256,
|
| 23 |
+
"relative_attention": true,
|
| 24 |
+
"torch_dtype": "float32",
|
| 25 |
+
"transformers_version": "4.21.0",
|
| 26 |
+
"type_vocab_size": 0,
|
| 27 |
+
"vocab_size": 128100,
|
| 28 |
+
"target_sizes": {
|
| 29 |
+
"task_type": 8,
|
| 30 |
+
"creativity_scope": 5,
|
| 31 |
+
"reasoning": 5,
|
| 32 |
+
"contextual_knowledge": 5,
|
| 33 |
+
"number_of_few_shots": 5,
|
| 34 |
+
"domain_knowledge": 5,
|
| 35 |
+
"no_label_reason": 5,
|
| 36 |
+
"constraint_ct": 5
|
| 37 |
+
},
|
| 38 |
+
"task_type_map": {
|
| 39 |
+
"0": "Open QA",
|
| 40 |
+
"1": "Closed QA",
|
| 41 |
+
"2": "Summarization",
|
| 42 |
+
"3": "Text Generation",
|
| 43 |
+
"4": "Code Generation",
|
| 44 |
+
"5": "Chatbot",
|
| 45 |
+
"6": "Classification",
|
| 46 |
+
"7": "Rewrite",
|
| 47 |
+
"8": "Brainstorming",
|
| 48 |
+
"9": "Extraction",
|
| 49 |
+
"10": "Other"
|
| 50 |
+
},
|
| 51 |
+
"weights_map": {
|
| 52 |
+
"creativity_scope": [0.0, 0.25, 0.5, 0.75, 1.0],
|
| 53 |
+
"reasoning": [0.0, 0.25, 0.5, 0.75, 1.0],
|
| 54 |
+
"contextual_knowledge": [0.0, 0.25, 0.5, 0.75, 1.0],
|
| 55 |
+
"number_of_few_shots": [0.0, 1.0, 2.0, 3.0, 4.0],
|
| 56 |
+
"domain_knowledge": [0.0, 0.25, 0.5, 0.75, 1.0],
|
| 57 |
+
"no_label_reason": [0.0, 0.25, 0.5, 0.75, 1.0],
|
| 58 |
+
"constraint_ct": [0.0, 0.25, 0.5, 0.75, 1.0]
|
| 59 |
+
},
|
| 60 |
+
"divisor_map": {
|
| 61 |
+
"creativity_scope": 1.0,
|
| 62 |
+
"reasoning": 1.0,
|
| 63 |
+
"contextual_knowledge": 1.0,
|
| 64 |
+
"number_of_few_shots": 1.0,
|
| 65 |
+
"domain_knowledge": 1.0,
|
| 66 |
+
"no_label_reason": 1.0,
|
| 67 |
+
"constraint_ct": 1.0
|
| 68 |
+
},
|
| 69 |
+
"quantized": true,
|
| 70 |
+
"quantization_method": "dynamic",
|
| 71 |
+
"framework": "onnx",
|
| 72 |
+
"optimized_for": "cpu",
|
| 73 |
+
"file_name": "model_quantized.onnx",
|
| 74 |
+
"quantization_config": {
|
| 75 |
+
"format": "QOperator",
|
| 76 |
+
"mode": "IntegerOps",
|
| 77 |
+
"activations_dtype": "QUInt8",
|
| 78 |
+
"weights_dtype": "QInt8",
|
| 79 |
+
"is_static": false
|
| 80 |
+
}
|
| 81 |
+
}
|