Training in progress, epoch 4

config.json

@@ -1,12 +1,12 @@
 {
-  "_name_or_path": "google/mt5-small",
+  "_name_or_path": "google/mt5-base",
   "architectures": [
     "MT5ForConditionalGeneration"
   ],
   "classifier_dropout": 0.0,
-  "d_ff": 1024,
+  "d_ff": 2048,
   "d_kv": 64,
-  "d_model": 512,
+  "d_model": 768,
   "decoder_start_token_id": 0,
   "dense_act_fn": "gelu_new",
   "dropout_rate": 0.1,
@@ -17,9 +17,10 @@
   "is_gated_act": true,
   "layer_norm_epsilon": 1e-06,
   "model_type": "mt5",
-  "num_decoder_layers": 8,
-  "num_heads": 6,
-  "num_layers": 8,
+  "num_decoder_layers": 12,
+  "num_heads": 12,
+  "num_layers": 12,
+  "output_past": true,
   "pad_token_id": 0,
   "relative_attention_max_distance": 128,
   "relative_attention_num_buckets": 32,

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c6dc19973c8451f13ecd46c12c2e3c448cb66c7a4cc303b6a56c1b417983b12a
-size 1200729512
+oid sha256:9d2aaca3d6c7e880e5babba51d8e1093a502645168a43bcb2fae5e048a3bd380
+size 2329638768

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5d7acd8dda6ecb9007309f9db8780088a2afc9619661e047ee6ca314374812ed
+oid sha256:d479a3ac8d594b2e96048ffa2a2c2ee3fbe9942a09fda4f363e2fd050344e301
 size 5176