Training in progress, step 500

config.json

@@ -1,12 +1,12 @@
 {
-  "_name_or_path": "csebuetnlp/mT5_multilingual_XLSum",
+  "_name_or_path": "google/mt5-small",
   "architectures": [
     "MT5ForConditionalGeneration"
   ],
   "classifier_dropout": 0.0,
-  "d_ff": 2048,
+  "d_ff": 1024,
   "d_kv": 64,
-  "d_model": 768,
+  "d_model": 512,
   "decoder_start_token_id": 0,
   "dense_act_fn": "gelu_new",
   "dropout_rate": 0.1,
@@ -16,15 +16,10 @@
   "is_encoder_decoder": true,
   "is_gated_act": true,
   "layer_norm_epsilon": 1e-06,
-  "length_penalty": 0.6,
-  "max_length": 84,
   "model_type": "mt5",
-  "no_repeat_ngram_size": 2,
-  "num_beams": 4,
-  "num_decoder_layers": 12,
-  "num_heads": 12,
-  "num_layers": 12,
-  "output_past": true,
+  "num_decoder_layers": 8,
+  "num_heads": 6,
+  "num_layers": 8,
   "pad_token_id": 0,
   "relative_attention_max_distance": 128,
   "relative_attention_num_buckets": 32,

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:26ef9fdd85c0a3334c2e2f24ce09764151902f6ccafa267952474fe63ff3a911
-size 2329702581
+oid sha256:30475f23246f343811a4a98ede1e7326574a5b3b952c9bb7146332f7dedab48d
+size 1200772613

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c085e5f6c8c6ab6fd62899af0165d6828dcbb93d774696bb3ee40897a2812ec5
+oid sha256:a4a0c49f89ea96332a9828c572b00ed6fd80c0a5caf256fdf438153f38c68cc3
 size 4219