Upload BartForConditionalGeneration

config.json

@@ -1,6 +1,9 @@
 {
+  "_num_labels": 3,
   "activation_dropout": 0.0,
   "activation_function": "gelu_new",
+  "add_bias_logits": false,
+  "add_final_layer_norm": false,
   "architectures": [
     "BartForConditionalGeneration"
   ],
@@ -8,12 +11,14 @@
   "bos_token_id": 2,
   "classifier_dropout": 0.0,
   "d_model": 768,
-  "decoder_attention_heads": 12,
+  "decoder_attention_heads": 16,
   "decoder_ffn_dim": 4096,
   "decoder_layerdrop": 0.0,
   "decoder_layers": 28,
   "decoder_start_token_id": 2,
+  "do_blenderbot_90_layernorm": false,
   "dropout": 0.1,
+  "early_stopping": true,
   "encoder_attention_heads": 24,
   "encoder_ffn_dim": 4096,
   "encoder_layerdrop": 0.0,
@@ -32,9 +37,12 @@
     "LABEL_1": 1,
     "LABEL_2": 2
   },
+  "max_length": 256,
   "max_position_embeddings": 1026,
+  "min_length": 20,
   "model_type": "bart",
-  "num_hidden_layers": 2,
+  "no_repeat_ngram_size": 3,
+  "num_hidden_layers": 9,
   "pad_token_id": 3,
   "scale_embedding": false,
   "torch_dtype": "float32",

generation_config.json

@@ -2,8 +2,12 @@
   "_from_model_config": true,
   "bos_token_id": 2,
   "decoder_start_token_id": 2,
+  "early_stopping": true,
   "eos_token_id": 2,
   "forced_eos_token_id": 2,
+  "max_length": 256,
+  "min_length": 20,
+  "no_repeat_ngram_size": 3,
   "pad_token_id": 3,
   "transformers_version": "4.26.0"
 }

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:25ddf17bc587aa7903d03a4d2418b9949bd985b3b48ca16f7ba14f086828b0b4
+oid sha256:9c4993e36a808ec20a59e2e1c3eb84bba2c8458877d9113ad5af1397b975991e
 size 1645594701