End of training

README.md

@@ -1,5 +1,6 @@
 ---
-base_model: eenzeenee/t5-small-korean-summarization
+license: apache-2.0
+base_model: t5-small
 tags:
 - generated_from_trainer
 model-index:
@@ -12,9 +13,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 # FingerFriend-t5-small
 
-This model is a fine-tuned version of [eenzeenee/t5-small-korean-summarization](https://huggingface.co/eenzeenee/t5-small-korean-summarization) on the None dataset.
+This model is a fine-tuned version of [t5-small](https://huggingface.co/t5-small) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.6002
+- Loss: 0.7464
 
 ## Model description
 
@@ -34,8 +35,8 @@ More information needed
 
 The following hyperparameters were used during training:
 - learning_rate: 2e-05
-- train_batch_size: 8
-- eval_batch_size: 8
+- train_batch_size: 16
+- eval_batch_size: 16
 - seed: 42
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
@@ -45,12 +46,26 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch | Step | Validation Loss |
 |:-------------:|:-----:|:----:|:---------------:|
-| 1.4509        | 1.0   | 342  | 0.7916          |
-| 0.8917        | 2.0   | 684  | 0.7062          |
-| 0.7846        | 3.0   | 1026 | 0.6640          |
-| 0.72          | 4.0   | 1368 | 0.6334          |
-| 0.6804        | 5.0   | 1710 | 0.6100          |
-| 0.6387        | 6.0   | 2052 | 0.6002          |
+| 1.6293        | 1.0   | 171  | 1.1671          |
+| 1.195         | 2.0   | 342  | 1.0246          |
+| 1.085         | 3.0   | 513  | 0.9553          |
+| 1.0207        | 4.0   | 684  | 0.9096          |
+| 0.9631        | 5.0   | 855  | 0.8782          |
+| 0.9283        | 6.0   | 1026 | 0.8445          |
+| 0.8987        | 7.0   | 1197 | 0.8352          |
+| 0.8716        | 8.0   | 1368 | 0.8123          |
+| 0.8556        | 9.0   | 1539 | 0.7983          |
+| 0.8375        | 10.0  | 1710 | 0.7923          |
+| 0.8239        | 11.0  | 1881 | 0.7757          |
+| 0.8184        | 12.0  | 2052 | 0.7716          |
+| 0.8053        | 13.0  | 2223 | 0.7642          |
+| 0.7929        | 14.0  | 2394 | 0.7647          |
+| 0.7867        | 15.0  | 2565 | 0.7597          |
+| 0.7817        | 16.0  | 2736 | 0.7529          |
+| 0.7751        | 17.0  | 2907 | 0.7506          |
+| 0.7705        | 18.0  | 3078 | 0.7472          |
+| 0.7657        | 19.0  | 3249 | 0.7467          |
+| 0.7665        | 20.0  | 3420 | 0.7464          |
 
 
 ### Framework versions

config.json

@@ -1,33 +1,61 @@
 {
-  "_name_or_path": "eenzeenee/t5-small-korean-summarization",
+  "_name_or_path": "t5-small",
   "architectures": [
     "T5ForConditionalGeneration"
   ],
   "classifier_dropout": 0.0,
-  "d_ff": 1024,
+  "d_ff": 2048,
   "d_kv": 64,
   "d_model": 512,
   "decoder_start_token_id": 0,
-  "dense_act_fn": "gelu_new",
+  "dense_act_fn": "relu",
   "dropout_rate": 0.1,
   "eos_token_id": 1,
-  "feed_forward_proj": "gated-gelu",
+  "feed_forward_proj": "relu",
   "initializer_factor": 1.0,
   "is_encoder_decoder": true,
-  "is_gated_act": true,
+  "is_gated_act": false,
   "layer_norm_epsilon": 1e-06,
-  "max_length": 128,
   "model_type": "t5",
-  "num_decoder_layers": 8,
-  "num_heads": 6,
-  "num_layers": 8,
+  "n_positions": 512,
+  "num_decoder_layers": 6,
+  "num_heads": 8,
+  "num_layers": 6,
   "output_past": true,
   "pad_token_id": 0,
   "relative_attention_max_distance": 128,
   "relative_attention_num_buckets": 32,
-  "tie_word_embeddings": false,
+  "task_specific_params": {
+    "summarization": {
+      "early_stopping": true,
+      "length_penalty": 2.0,
+      "max_length": 200,
+      "min_length": 30,
+      "no_repeat_ngram_size": 3,
+      "num_beams": 4,
+      "prefix": "summarize: "
+    },
+    "translation_en_to_de": {
+      "early_stopping": true,
+      "max_length": 300,
+      "num_beams": 4,
+      "prefix": "translate English to German: "
+    },
+    "translation_en_to_fr": {
+      "early_stopping": true,
+      "max_length": 300,
+      "num_beams": 4,
+      "prefix": "translate English to French: "
+    },
+    "translation_en_to_ro": {
+      "early_stopping": true,
+      "max_length": 300,
+      "num_beams": 4,
+      "prefix": "translate English to Romanian: "
+    }
+  },
   "torch_dtype": "float32",
   "transformers_version": "4.33.2",
   "use_cache": true,
-  "vocab_size": 50358
+  "vocab_size": 32128
 }

generation_config.json

@@ -2,7 +2,6 @@
   "_from_model_config": true,
   "decoder_start_token_id": 0,
   "eos_token_id": 1,
-  "max_length": 128,
   "pad_token_id": 0,
   "transformers_version": "4.33.2"
 }

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:408db3f7ac52178898e790a6cd185b6a9e21de5a085d64c1af7778130ccf1a26
-size 382580229
+oid sha256:5ff34878fa65c518899de5ac2b5118ae03d3c7e5b607468518705daaddaa81a8
+size 242071641

special_tokens_map.json

@@ -103,5 +103,5 @@
   ],
   "eos_token": "</s>",
   "pad_token": "<pad>",
-  "unk_token": "<pad>"
+  "unk_token": "<unk>"
 }

tokenizer_config.json

@@ -104,12 +104,8 @@
   "clean_up_tokenization_spaces": true,
   "eos_token": "</s>",
   "extra_ids": 100,
-  "max_length": 128,
-  "model_max_length": 128,
+  "model_max_length": 512,
   "pad_token": "<pad>",
-  "stride": 128,
   "tokenizer_class": "T5Tokenizer",
-  "truncation_side": "right",
-  "truncation_strategy": "longest_first",
-  "unk_token": "<pad>"
+  "unk_token": "<unk>"
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c0383c3fd8f425030cfa2c0c23ae01db1ffb7eed00a42904bef85495ec164cff
+oid sha256:1292f1a8edea159be8f39c77046e007078f23fed971ae55ff1d661607c577d38
 size 4155