Training in progress, step 500

config.json

@@ -1,11 +1,11 @@
 {
-  "_name_or_path": "t5-large",
+  "_name_or_path": "t5-small",
   "architectures": [
     "T5ForConditionalGeneration"
   ],
-  "d_ff": 4096,
+  "d_ff": 2048,
   "d_kv": 64,
-  "d_model": 1024,
+  "d_model": 512,
   "decoder_start_token_id": 0,
   "dense_act_fn": "relu",
   "dropout_rate": 0.1,
@@ -17,9 +17,9 @@
   "layer_norm_epsilon": 1e-06,
   "model_type": "t5",
   "n_positions": 512,
-  "num_decoder_layers": 24,
-  "num_heads": 16,
-  "num_layers": 24,
+  "num_decoder_layers": 6,
+  "num_heads": 8,
+  "num_layers": 6,
   "output_past": true,
   "pad_token_id": 0,
   "relative_attention_max_distance": 128,

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f57995ad781ab482844f8a340368a83010d743c5e045046fe97213b6aed40283
-size 2950848513
+oid sha256:d64a329f7dab923061dcee65b7c7e7f1502fca8381a9ef43d6231695e32edb0f
+size 242071641

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:84b5795efa7613f5a39921f6527f60a83386f9e91df1f2185993c5a48edb6b97
+oid sha256:e9a65a217ffc90d91d6f7eead9a3745ec63feb75c58be61fc8fb26606d0c725e
 size 4091