Update config.json
Browse files- config.json +1 -31
config.json
CHANGED
|
@@ -1,48 +1,23 @@
|
|
| 1 |
{
|
| 2 |
-
"_num_labels": 2,
|
| 3 |
"architectures": [
|
| 4 |
"T5ForConditionalGeneration"
|
| 5 |
],
|
| 6 |
-
"bos_token_id": null,
|
| 7 |
"d_ff": 256,
|
| 8 |
"d_kv": 8,
|
| 9 |
"d_model": 64,
|
| 10 |
"decoder_start_token_id": 0,
|
| 11 |
-
"do_sample": false,
|
| 12 |
"dropout_rate": 0.1,
|
| 13 |
-
"early_stopping": false,
|
| 14 |
"eos_token_id": 1,
|
| 15 |
-
"finetuning_task": null,
|
| 16 |
-
"id2label": {
|
| 17 |
-
"0": "LABEL_0",
|
| 18 |
-
"1": "LABEL_1"
|
| 19 |
-
},
|
| 20 |
"initializer_factor": 1.0,
|
| 21 |
-
"is_decoder": false,
|
| 22 |
"is_encoder_decoder": true,
|
| 23 |
-
"label2id": {
|
| 24 |
-
"LABEL_0": 0,
|
| 25 |
-
"LABEL_1": 1
|
| 26 |
-
},
|
| 27 |
"layer_norm_epsilon": 1e-06,
|
| 28 |
-
"length_penalty": 1.0,
|
| 29 |
-
"max_length": 20,
|
| 30 |
-
"min_length": 0,
|
| 31 |
"model_type": "t5",
|
| 32 |
"n_positions": 512,
|
| 33 |
-
"no_repeat_ngram_size": 0,
|
| 34 |
-
"num_beams": 1,
|
| 35 |
"num_heads": 2,
|
| 36 |
"num_layers": 2,
|
| 37 |
-
"num_return_sequences": 1,
|
| 38 |
-
"output_attentions": false,
|
| 39 |
-
"output_hidden_states": false,
|
| 40 |
"output_past": true,
|
| 41 |
"pad_token_id": 0,
|
| 42 |
-
"prefix": null,
|
| 43 |
-
"pruned_heads": {},
|
| 44 |
"relative_attention_num_buckets": 32,
|
| 45 |
-
"repetition_penalty": 1.0,
|
| 46 |
"task_specific_params": {
|
| 47 |
"summarization": {
|
| 48 |
"early_stopping": true,
|
|
@@ -72,10 +47,5 @@
|
|
| 72 |
"prefix": "translate English to Romanian: "
|
| 73 |
}
|
| 74 |
},
|
| 75 |
-
"temperature": 1.0,
|
| 76 |
-
"top_k": 50,
|
| 77 |
-
"top_p": 1.0,
|
| 78 |
-
"torchscript": false,
|
| 79 |
-
"use_bfloat16": false,
|
| 80 |
"vocab_size": 32128
|
| 81 |
-
}
|
| 1 |
{
|
|
|
|
| 2 |
"architectures": [
|
| 3 |
"T5ForConditionalGeneration"
|
| 4 |
],
|
|
|
|
| 5 |
"d_ff": 256,
|
| 6 |
"d_kv": 8,
|
| 7 |
"d_model": 64,
|
| 8 |
"decoder_start_token_id": 0,
|
|
|
|
| 9 |
"dropout_rate": 0.1,
|
|
|
|
| 10 |
"eos_token_id": 1,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 11 |
"initializer_factor": 1.0,
|
|
|
|
| 12 |
"is_encoder_decoder": true,
|
|
|
|
|
|
|
|
|
|
|
|
|
| 13 |
"layer_norm_epsilon": 1e-06,
|
|
|
|
|
|
|
|
|
|
| 14 |
"model_type": "t5",
|
| 15 |
"n_positions": 512,
|
|
|
|
|
|
|
| 16 |
"num_heads": 2,
|
| 17 |
"num_layers": 2,
|
|
|
|
|
|
|
|
|
|
| 18 |
"output_past": true,
|
| 19 |
"pad_token_id": 0,
|
|
|
|
|
|
|
| 20 |
"relative_attention_num_buckets": 32,
|
|
|
|
| 21 |
"task_specific_params": {
|
| 22 |
"summarization": {
|
| 23 |
"early_stopping": true,
|
| 47 |
"prefix": "translate English to Romanian: "
|
| 48 |
}
|
| 49 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 50 |
"vocab_size": 32128
|
| 51 |
+
}
|