Commit
•
622c63f
1
Parent(s):
19ad2da
Upload config.json with huggingface_hub
Browse files- config.json +60 -0
config.json
ADDED
|
@@ -0,0 +1,60 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alibi_mode": "symetric",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"FlashT5ForConditionalGeneration"
|
| 5 |
+
],
|
| 6 |
+
"attention_dropout_rate": 0.0,
|
| 7 |
+
"attention_scale": 1.0,
|
| 8 |
+
"attention_type": "ref",
|
| 9 |
+
"auto_map": {
|
| 10 |
+
"AutoConfig": "configuration_flash_t5.FlashT5Config",
|
| 11 |
+
"AutoModel": "modeling_flash_t5.FlashT5ForConditionalGeneration",
|
| 12 |
+
"AutoModelForQuestionAnswering": "custom_heads_flash_t5.FlashT5ForQuestionAnswering",
|
| 13 |
+
"AutoModelForSeq2SeqLM": "modeling_flash_t5.FlashT5ForConditionalGeneration",
|
| 14 |
+
"AutoModelForSequenceClassification": "custom_heads_flash_t5.FlashT5ForSequenceClassification",
|
| 15 |
+
"AutoModelForTokenClassification": "custom_heads_flash_t5.FlashT5ForTokenClassification"
|
| 16 |
+
},
|
| 17 |
+
"classifier_dropout": 0.0,
|
| 18 |
+
"d_ff": 2816,
|
| 19 |
+
"d_kv": 64,
|
| 20 |
+
"d_model": 1024,
|
| 21 |
+
"decoder_start_token_id": 0,
|
| 22 |
+
"dense_act_fn": "relu",
|
| 23 |
+
"dropout_rate": 0.0,
|
| 24 |
+
"eos_token_id": 1,
|
| 25 |
+
"feed_forward_proj": "relu",
|
| 26 |
+
"fire_mlp_width": 32,
|
| 27 |
+
"initializer_factor": 1.0,
|
| 28 |
+
"is_encoder_decoder": false,
|
| 29 |
+
"is_gated_act": false,
|
| 30 |
+
"label_smoothing": 0.1,
|
| 31 |
+
"layer_norm_epsilon": 1e-06,
|
| 32 |
+
"max_sequence_length": 1024,
|
| 33 |
+
"model_type": "flash_t5",
|
| 34 |
+
"num_decoder_layers": 24,
|
| 35 |
+
"num_heads": 16,
|
| 36 |
+
"num_layers": 24,
|
| 37 |
+
"pad_token_id": 0,
|
| 38 |
+
"position_encoding_type": "t5",
|
| 39 |
+
"relative_attention_max_distance": 128,
|
| 40 |
+
"relative_attention_num_buckets": 32,
|
| 41 |
+
"rotary_base": 10000,
|
| 42 |
+
"rotary_emb_fraction": 1.0,
|
| 43 |
+
"rotary_interleaved": false,
|
| 44 |
+
"rotary_scale_base": null,
|
| 45 |
+
"tie_word_embeddings": false,
|
| 46 |
+
"torch_dtype": "float32",
|
| 47 |
+
"transformers_version": "4.39.3",
|
| 48 |
+
"use_cache": true,
|
| 49 |
+
"use_flash_attention": "triton",
|
| 50 |
+
"use_full_bias_size": false,
|
| 51 |
+
"use_gelu_act": true,
|
| 52 |
+
"use_glu_mlp": true,
|
| 53 |
+
"use_masking": false,
|
| 54 |
+
"use_randomized_position_encoding": false,
|
| 55 |
+
"use_triton_crossentropy": false,
|
| 56 |
+
"use_triton_gated_mlp": false,
|
| 57 |
+
"use_triton_layernorm": false,
|
| 58 |
+
"vocab_size": 32128,
|
| 59 |
+
"z_loss": 0.0001
|
| 60 |
+
}
|