Training in progress, step 500

config.json

@@ -1,12 +1,12 @@
 {
-  "_name_or_path": "badokorach/flan-t5-small-qa",
+  "_name_or_path": "google/flan-t5-base",
   "architectures": [
     "T5ForConditionalGeneration"
   ],
   "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,
@@ -18,9 +18,9 @@
   "layer_norm_epsilon": 1e-06,
   "model_type": "t5",
   "n_positions": 512,
-  "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,
@@ -56,7 +56,7 @@
   },
   "tie_word_embeddings": false,
   "torch_dtype": "float32",
-  "transformers_version": "4.33.2",
+  "transformers_version": "4.33.3",
   "use_cache": true,
   "vocab_size": 32128
 }

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5425875613158ac52b2e86bfc03f6821d0ce76289be2ec0c87d680613cbe5afe
-size 307910149
+oid sha256:1ee89397921311272a2b393f558b82940ae66da399c752a46bcdd0094e3575c6
+size 990408885

tokenizer_config.json

@@ -104,16 +104,9 @@
   "clean_up_tokenization_spaces": true,
   "eos_token": "</s>",
   "extra_ids": 100,
-  "max_length": 512,
   "model_max_length": 512,
-  "pad_to_multiple_of": null,
   "pad_token": "<pad>",
-  "pad_token_type_id": 0,
-  "padding_side": "right",
   "sp_model_kwargs": {},
-  "stride": 0,
   "tokenizer_class": "T5Tokenizer",
-  "truncation_side": "right",
-  "truncation_strategy": "longest_first",
   "unk_token": "<unk>"
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:af82bc35b06a23db69f1e50063f7047ad91033415cab8ce876168bbfa609e4f3
+oid sha256:1d6935f00d2d0434138a15b530efc43fdc0cb0edbd152267f1f32314f0712f74
 size 4027