Training in progress, step 500

config.json

@@ -1,12 +1,12 @@
 {
-  "_name_or_path": "google/mt5-small",
+  "_name_or_path": "google/mt5-base",
   "architectures": [
     "MT5ForTokenClassification"
   ],
   "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,
@@ -35,16 +35,17 @@
   },
   "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,
   "tie_word_embeddings": false,
   "tokenizer_class": "T5Tokenizer",
   "torch_dtype": "float32",
-  "transformers_version": "4.41.2",
+  "transformers_version": "4.39.3",
   "use_cache": true,
   "vocab_size": 250112
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:231687f148d897a59a4d8f2bdd776f99c893052f03278ec4295fa01a1df1f3a3
-size 587786644
+oid sha256:2ae0cea7187e3173da2386c7bbff891bdd61e8e63d72bbbdabdcc48655874f79
+size 1108197244

runs/Jul17_16-14-57_izaitova-23814.0-cl18lx.lsv.uni-saarland.de/events.out.tfevents.1721232901.izaitova-23814.0-cl18lx.lsv.uni-saarland.de.1580.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ee2429ba3a832e8612b5778a590a845a11ea08a4cfece34e4d4b000971d874fc
+size 6530

tokenizer.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ddb1b00035edffde435e24e69011b7f552691405237455cb7c3a59c890682807
-size 16330540
+oid sha256:87a036c0dfd2d80e1202a7e2961aeee653ff63d67cd369b155c78a6e2003a390
+size 16330562

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3d167e0d0c84c62c16b1e5e8bd8642cf61722e46ae06a2f7feba3a45c42e4acf
-size 5112
+oid sha256:2104188fa8b5ec7891060cbe2cc12d6d3c603b0fe2b5af25df130abd5c27d8c8
+size 4527