Trained on 20min [Backtranslation Decoder-Dropout 0.1]

config.json

@@ -18,7 +18,7 @@
   "decoder_layerdrop": 0.0,
   "decoder_layers": 12,
   "decoder_start_token_id": 250003,
-  "dropout": 0.0,
+  "dropout": 0.1,
   "encoder_attention_heads": 16,
   "encoder_ffn_dim": 4096,
   "encoder_layerdrop": 0.0,
@@ -55,7 +55,7 @@
     }
   },
   "torch_dtype": "float32",
-  "transformers_version": "4.25.1",
+  "transformers_version": "4.27.4",
   "use_cache": true,
   "vocab_size": 250027
 }

generation_config.json

@@ -0,0 +1,11 @@
+{
+  "bos_token_id": 0,
+  "decoder_start_token_id": 250003,
+  "eos_token_id": 2,
+  "forced_bos_token_id": 0,
+  "forced_eos_token_id": 2,
+  "max_length": 1024,
+  "num_beams": 5,
+  "pad_token_id": 1,
+  "transformers_version": "4.27.4"
+}

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:429e5eac36f52e6ed11a17add62d2f384e5b3a78038b3d494ca2e802bd565606
+oid sha256:e66cb02577b74ca6a8d464d407907813b367361e2768c651c2a9f057536e5e02
 size 2444583325