Training in progress, step 336

config.json

@@ -3,9 +3,9 @@
     "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,
@@ -16,9 +16,10 @@
   "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 sha256:8b64eaa4c524d70a35bade3e3225c9a841c0a95c4b13d72a1aec31fc4b9a9dcb
-size 1200729512
+oid sha256:eae042c2b6232537c6c487deb34f7cea8caf142d3069b3b460607097359bab67
+size 2329638768

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:07017af49eafac0fe195264b24368b06d35fb983d19e0c55ada80a21de831bad
+oid sha256:f2e9347e332ad304a8dc6cf9f8abd18d6784a3f9821e57c43da01ecba1da096e
 size 5432