End of training

README.md

@@ -1,8 +1,10 @@
 ---
 license: apache-2.0
-base_model: google/flan-t5-base
+base_model: t5-base
 tags:
 - generated_from_trainer
+metrics:
+- rouge
 model-index:
 - name: ingredient_prune
   results: []
@@ -13,9 +15,14 @@ should probably proofread and complete it, then remove this comment. -->
 
 # ingredient_prune
 
-This model is a fine-tuned version of [google/flan-t5-base](https://huggingface.co/google/flan-t5-base) on the None dataset.
+This model is a fine-tuned version of [t5-base](https://huggingface.co/t5-base) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.0140
+- Loss: 0.0255
+- Rouge1: 88.3061
+- Rouge2: 76.6099
+- Rougel: 88.3242
+- Rougelsum: 88.2429
+- Gen Len: 10.5872
 
 ## Model description
 
@@ -35,42 +42,72 @@ 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: 4
+- eval_batch_size: 4
 - seed: 42
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
-- num_epochs: 20
+- num_epochs: 5
 - mixed_precision_training: Native AMP
 
 ### Training results
 
-| Training Loss | Epoch | Step | Validation Loss |
-|:-------------:|:-----:|:----:|:---------------:|
-| 17.212        | 0.82  | 100  | 4.0184          |
-| 2.4003        | 1.64  | 200  | 0.2977          |
-| 0.3173        | 2.46  | 300  | 0.0432          |
-| 0.068         | 3.28  | 400  | 0.0203          |
-| 0.0356        | 4.1   | 500  | 0.0167          |
-| 0.026         | 4.92  | 600  | 0.0145          |
-| 0.0194        | 5.74  | 700  | 0.0138          |
-| 0.0179        | 6.56  | 800  | 0.0132          |
-| 0.0154        | 7.38  | 900  | 0.0134          |
-| 0.0133        | 8.2   | 1000 | 0.0134          |
-| 0.0125        | 9.02  | 1100 | 0.0132          |
-| 0.013         | 9.84  | 1200 | 0.0131          |
-| 0.011         | 10.66 | 1300 | 0.0136          |
-| 0.0106        | 11.48 | 1400 | 0.0133          |
-| 0.0098        | 12.3  | 1500 | 0.0133          |
-| 0.0093        | 13.11 | 1600 | 0.0134          |
-| 0.009         | 13.93 | 1700 | 0.0138          |
-| 0.0085        | 14.75 | 1800 | 0.0141          |
-| 0.0083        | 15.57 | 1900 | 0.0136          |
-| 0.0083        | 16.39 | 2000 | 0.0139          |
-| 0.0082        | 17.21 | 2100 | 0.0139          |
-| 0.0083        | 18.03 | 2200 | 0.0139          |
-| 0.0076        | 18.85 | 2300 | 0.0140          |
-| 0.0085        | 19.67 | 2400 | 0.0140          |
+| Training Loss | Epoch | Step | Validation Loss | Rouge1  | Rouge2  | Rougel  | Rougelsum | Gen Len |
+|:-------------:|:-----:|:----:|:---------------:|:-------:|:-------:|:-------:|:---------:|:-------:|
+| 2.9499        | 0.09  | 10   | 1.3100          | 33.1645 | 23.9561 | 32.6647 | 32.7137   | 14.7431 |
+| 1.9454        | 0.18  | 20   | 0.6787          | 30.1119 | 21.203  | 29.5079 | 29.6061   | 13.8349 |
+| 1.309         | 0.28  | 30   | 0.5147          | 25.3399 | 17.694  | 24.4102 | 24.4425   | 11.6514 |
+| 1.0307        | 0.37  | 40   | 0.4398          | 17.4522 | 11.66   | 16.2846 | 16.3817   | 8.5413  |
+| 0.9574        | 0.46  | 50   | 0.4302          | 16.6745 | 10.6799 | 15.8568 | 16.4301   | 8.0092  |
+| 0.7183        | 0.55  | 60   | 0.3818          | 14.4343 | 9.4646  | 13.9825 | 14.1979   | 6.9725  |
+| 0.5636        | 0.64  | 70   | 0.3096          | 9.4156  | 5.2844  | 9.0143  | 9.239     | 5.5596  |
+| 0.4603        | 0.73  | 80   | 0.2664          | 8.6106  | 4.7574  | 7.9285  | 8.4429    | 5.0917  |
+| 0.4607        | 0.83  | 90   | 0.2319          | 6.7868  | 3.9309  | 6.1844  | 6.7007    | 3.8349  |
+| 0.352         | 0.92  | 100  | 0.1991          | 6.2965  | 3.5572  | 5.3616  | 5.9941    | 3.2661  |
+| 0.3426        | 1.01  | 110  | 0.1735          | 6.1795  | 3.1174  | 5.3783  | 5.9261    | 3.3119  |
+| 0.2901        | 1.1   | 120  | 0.1553          | 5.5031  | 2.739   | 4.9926  | 5.5079    | 3.1376  |
+| 0.3619        | 1.19  | 130  | 0.1452          | 4.1403  | 1.8462  | 4.0877  | 4.1877    | 3.0092  |
+| 0.2509        | 1.28  | 140  | 0.1338          | 4.1399  | 1.8019  | 3.9836  | 4.1506    | 2.9541  |
+| 0.1938        | 1.38  | 150  | 0.1187          | 2.9515  | 1.2174  | 2.7845  | 3.0192    | 2.2569  |
+| 0.1987        | 1.47  | 160  | 0.1068          | 4.8991  | 3.4459  | 4.7552  | 4.9489    | 2.1284  |
+| 0.1702        | 1.56  | 170  | 0.0983          | 8.7082  | 5.5788  | 8.5531  | 8.8267    | 3.4587  |
+| 0.1535        | 1.65  | 180  | 0.0871          | 11.5572 | 7.6669  | 11.4688 | 11.5381   | 4.6972  |
+| 0.1629        | 1.74  | 190  | 0.0771          | 16.33   | 11.587  | 16.0842 | 16.1965   | 6.6055  |
+| 0.1618        | 1.83  | 200  | 0.0690          | 21.4186 | 14.9296 | 21.2789 | 21.2002   | 8.367   |
+| 0.1617        | 1.93  | 210  | 0.0628          | 27.6198 | 19.8907 | 27.4479 | 27.4515   | 10.3394 |
+| 0.1136        | 2.02  | 220  | 0.0572          | 36.7416 | 28.2903 | 36.7181 | 36.719    | 12.3578 |
+| 0.1278        | 2.11  | 230  | 0.0526          | 46.9007 | 36.6481 | 47.1002 | 46.8623   | 13.7064 |
+| 0.0915        | 2.2   | 240  | 0.0486          | 56.1238 | 45.5624 | 56.3372 | 56.0369   | 14.1101 |
+| 0.0736        | 2.29  | 250  | 0.0448          | 63.3857 | 51.8889 | 63.6163 | 63.2685   | 13.4771 |
+| 0.0855        | 2.39  | 260  | 0.0420          | 72.669  | 59.9359 | 72.7393 | 72.6055   | 12.3486 |
+| 0.0921        | 2.48  | 270  | 0.0388          | 78.2224 | 65.2581 | 78.2789 | 77.9532   | 11.3578 |
+| 0.077         | 2.57  | 280  | 0.0364          | 82.3881 | 68.397  | 82.4999 | 82.3175   | 10.5872 |
+| 0.0848        | 2.66  | 290  | 0.0347          | 85.4014 | 72.793  | 85.495  | 85.3917   | 10.633  |
+| 0.0978        | 2.75  | 300  | 0.0332          | 86.0947 | 72.9678 | 86.1325 | 86.0028   | 10.5138 |
+| 0.0635        | 2.84  | 310  | 0.0323          | 86.158  | 73.833  | 86.2727 | 86.1471   | 10.5596 |
+| 0.0555        | 2.94  | 320  | 0.0314          | 86.0306 | 73.8297 | 86.0421 | 85.9571   | 10.5688 |
+| 0.0792        | 3.03  | 330  | 0.0305          | 87.5066 | 75.3885 | 87.6496 | 87.3874   | 10.3761 |
+| 0.0536        | 3.12  | 340  | 0.0297          | 88.0844 | 75.8754 | 88.1956 | 87.9164   | 10.4954 |
+| 0.063         | 3.21  | 350  | 0.0290          | 88.0844 | 75.8754 | 88.1956 | 87.9164   | 10.4954 |
+| 0.0563        | 3.3   | 360  | 0.0283          | 88.0783 | 75.989  | 88.2233 | 87.9578   | 10.5138 |
+| 0.0547        | 3.39  | 370  | 0.0279          | 88.1265 | 76.3196 | 88.3078 | 88.0765   | 10.6147 |
+| 0.0635        | 3.49  | 380  | 0.0275          | 86.9846 | 74.8237 | 87.0556 | 86.9021   | 10.5872 |
+| 0.0835        | 3.58  | 390  | 0.0271          | 86.933  | 75.3277 | 87.0357 | 86.931    | 10.6147 |
+| 0.0628        | 3.67  | 400  | 0.0269          | 87.5981 | 75.5811 | 87.6905 | 87.4594   | 10.6789 |
+| 0.0554        | 3.76  | 410  | 0.0267          | 88.0124 | 76.5633 | 88.174  | 87.9292   | 10.578  |
+| 0.0342        | 3.85  | 420  | 0.0266          | 88.0124 | 76.5633 | 88.174  | 87.9292   | 10.578  |
+| 0.0396        | 3.94  | 430  | 0.0263          | 88.0064 | 76.6947 | 88.1712 | 87.9434   | 10.5872 |
+| 0.045         | 4.04  | 440  | 0.0262          | 87.7466 | 76.3605 | 87.8932 | 87.6273   | 10.5505 |
+| 0.0566        | 4.13  | 450  | 0.0262          | 87.8577 | 76.5633 | 88.0399 | 87.7835   | 10.6055 |
+| 0.0582        | 4.22  | 460  | 0.0261          | 87.8103 | 76.1351 | 87.9277 | 87.7032   | 10.6697 |
+| 0.051         | 4.31  | 470  | 0.0260          | 87.8103 | 76.1351 | 87.9277 | 87.7032   | 10.6697 |
+| 0.0398        | 4.4   | 480  | 0.0258          | 88.1974 | 76.4006 | 88.2158 | 88.0622   | 10.6789 |
+| 0.0364        | 4.5   | 490  | 0.0257          | 88.3353 | 76.5513 | 88.3291 | 88.2557   | 10.633  |
+| 0.0498        | 4.59  | 500  | 0.0257          | 88.4083 | 76.5513 | 88.4132 | 88.35     | 10.6147 |
+| 0.0406        | 4.68  | 510  | 0.0256          | 88.3061 | 76.6099 | 88.3242 | 88.2429   | 10.5872 |
+| 0.0403        | 4.77  | 520  | 0.0256          | 88.3061 | 76.6099 | 88.3242 | 88.2429   | 10.5872 |
+| 0.0421        | 4.86  | 530  | 0.0255          | 88.3061 | 76.6099 | 88.3242 | 88.2429   | 10.5872 |
+| 0.0271        | 4.95  | 540  | 0.0255          | 88.3061 | 76.6099 | 88.3242 | 88.2429   | 10.5872 |
 
 
 ### Framework versions

config.json

@@ -1,20 +1,20 @@
 {
-  "_name_or_path": "google/flan-t5-base",
+  "_name_or_path": "t5-base",
   "architectures": [
     "T5ForConditionalGeneration"
   ],
   "classifier_dropout": 0.0,
-  "d_ff": 2048,
+  "d_ff": 3072,
   "d_kv": 64,
   "d_model": 768,
   "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,
   "model_type": "t5",
   "n_positions": 512,
@@ -54,7 +54,6 @@
       "prefix": "translate English to Romanian: "
     }
   },
-  "tie_word_embeddings": false,
   "torch_dtype": "float32",
   "transformers_version": "4.38.2",
   "use_cache": true,

generation_config.json

@@ -1,5 +1,4 @@
 {
-  "_from_model_config": true,
   "decoder_start_token_id": 0,
   "eos_token_id": 1,
   "pad_token_id": 0,

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4a6d29efc5671f216a56027ac9f94db20161dd36fca529a1c2334b4f65cf718f
-size 990345064
+oid sha256:748d64e4c64aa8b8f55f19dedfee6ea3ef7c2a3a1c89678ea746ab2e5393f68e
+size 891644712

runs/Apr23_13-58-39_bcff8d0abbd9/events.out.tfevents.1713880729.bcff8d0abbd9.34.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3596b012ee2304b921f084307f4627d6bc65797e4617a22c98d7bfbe78ea646a
+size 17125

runs/Apr23_14-05-07_bcff8d0abbd9/events.out.tfevents.1713881108.bcff8d0abbd9.34.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ae48622f944c770ea51763277f7b1f5f83a5c9aa3e52d7add6df9a43aa2a8ebb
+size 5767

runs/Apr23_14-06-35_bcff8d0abbd9/events.out.tfevents.1713881196.bcff8d0abbd9.34.2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:71847a557d89aee0e4efad70c6311d1dc5b4115b23879d1d68bd70a11656be38
+size 52613

special_tokens_map.json

@@ -101,25 +101,7 @@
     "<extra_id_98>",
     "<extra_id_99>"
   ],
-  "eos_token": {
-    "content": "</s>",
-    "lstrip": false,
-    "normalized": false,
-    "rstrip": false,
-    "single_word": false
-  },
-  "pad_token": {
-    "content": "<pad>",
-    "lstrip": false,
-    "normalized": false,
-    "rstrip": false,
-    "single_word": false
-  },
-  "unk_token": {
-    "content": "<unk>",
-    "lstrip": false,
-    "normalized": false,
-    "rstrip": false,
-    "single_word": false
-  }
+  "eos_token": "</s>",
+  "pad_token": "<pad>",
+  "unk_token": "<unk>"
 }

tokenizer_config.json

@@ -932,7 +932,6 @@
   "extra_ids": 100,
   "model_max_length": 128,
   "pad_token": "<pad>",
-  "sp_model_kwargs": {},
   "tokenizer_class": "T5Tokenizer",
   "unk_token": "<unk>"
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:aaedefbc90eabca7075c2c56a07c9c692277ce37e1f1de92e6e0775bcb0e06dd
+oid sha256:ac460abe19f37f41bd8624e0c0f825129628eae8203e58c80e43fa839de25063
 size 5048