Upload 13 files

README.md

@@ -6,7 +6,7 @@ datasets:
 metrics:
 - rouge
 model-index:
-- name: small-2-2-t
+- name: small-2-2
   results:
   - task:
       name: Summarization
@@ -20,22 +20,22 @@ model-index:
     metrics:
     - name: Rouge1
       type: rouge
-      value: 26.601
+      value: 26.7124
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 should probably proofread and complete it, then remove this comment. -->
 
-# small-2-2-t
+# small-2-2
 
-This model is a fine-tuned version of [models/small-2-2](https://huggingface.co/models/small-2-2) on the xsum dataset.
+This model is a fine-tuned version of [x/small-2-2/](https://huggingface.co/x/small-2-2/) on the xsum dataset.
 It achieves the following results on the evaluation set:
-- Loss: 2.9093
-- Rouge1: 26.601
-- Rouge2: 7.1924
-- Rougel: 21.2586
-- Rougelsum: 21.2573
-- Gen Len: 26.3125
+- Loss: 2.9112
+- Rouge1: 26.7124
+- Rouge2: 7.3066
+- Rougel: 21.3783
+- Rougelsum: 21.3771
+- Gen Len: 26.3514
 
 ## Model description
 
@@ -55,10 +55,10 @@ More information needed
 
 The following hyperparameters were used during training:
 - learning_rate: 0.0001
-- train_batch_size: 32
-- eval_batch_size: 16
+- train_batch_size: 4
+- eval_batch_size: 4
 - seed: 42
-- gradient_accumulation_steps: 2
+- gradient_accumulation_steps: 16
 - total_train_batch_size: 64
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: constant
@@ -71,6 +71,6 @@ The following hyperparameters were used during training:
 ### Framework versions
 
 - Transformers 4.27.0.dev0
-- Pytorch 1.12.1+cu113
-- Datasets 2.10.0
-- Tokenizers 0.13.2
+- Pytorch 1.13.0+cu117
+- Datasets 2.7.1
+- Tokenizers 0.12.1

all_results.json

@@ -1,18 +1,18 @@
 {
     "epoch": 3.0,
-    "eval_gen_len": 26.312527588946764,
-    "eval_loss": 2.9093124866485596,
-    "eval_rouge1": 26.601,
-    "eval_rouge2": 7.1924,
-    "eval_rougeL": 21.2586,
-    "eval_rougeLsum": 21.2573,
-    "eval_runtime": 219.9179,
+    "eval_gen_len": 26.351372825990996,
+    "eval_loss": 2.9112093448638916,
+    "eval_rouge1": 26.7124,
+    "eval_rouge2": 7.3066,
+    "eval_rougeL": 21.3783,
+    "eval_rougeLsum": 21.3771,
+    "eval_runtime": 383.7315,
     "eval_samples": 11327,
-    "eval_samples_per_second": 51.506,
-    "eval_steps_per_second": 3.219,
-    "train_loss": 3.5344422363128087,
-    "train_runtime": 4207.4456,
+    "eval_samples_per_second": 29.518,
+    "eval_steps_per_second": 7.38,
+    "train_loss": 3.5328583012638624,
+    "train_runtime": 7182.8846,
     "train_samples": 204017,
-    "train_samples_per_second": 145.469,
-    "train_steps_per_second": 2.273
+    "train_samples_per_second": 85.21,
+    "train_steps_per_second": 1.331
 }

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "models/small-2-2",
+  "_name_or_path": "x/small-2-2/",
   "architectures": [
     "T5ForConditionalGeneration"
   ],

eval_results.json

@@ -1,13 +1,13 @@
 {
     "epoch": 3.0,
-    "eval_gen_len": 26.312527588946764,
-    "eval_loss": 2.9093124866485596,
-    "eval_rouge1": 26.601,
-    "eval_rouge2": 7.1924,
-    "eval_rougeL": 21.2586,
-    "eval_rougeLsum": 21.2573,
-    "eval_runtime": 219.9179,
+    "eval_gen_len": 26.351372825990996,
+    "eval_loss": 2.9112093448638916,
+    "eval_rouge1": 26.7124,
+    "eval_rouge2": 7.3066,
+    "eval_rougeL": 21.3783,
+    "eval_rougeLsum": 21.3771,
+    "eval_runtime": 383.7315,
     "eval_samples": 11327,
-    "eval_samples_per_second": 51.506,
-    "eval_steps_per_second": 3.219
+    "eval_samples_per_second": 29.518,
+    "eval_steps_per_second": 7.38
 }

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:38bf9caea956ad1781e4c9e2ad1af5ba1f8c629bc1388f9c6c4c36361fd50829
-size 175679497
+oid sha256:b144b5388dedfbfa6856b537a14d03e4a277eb349bc3107dc28b545282cb8a99
+size 175680397

spiece.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d60acb128cf7b7f2536e8f38a5b18a05535c9e14c7a355904270e15b0945ea86
+size 791656

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 3.0,
-    "train_loss": 3.5344422363128087,
-    "train_runtime": 4207.4456,
+    "train_loss": 3.5328583012638624,
+    "train_runtime": 7182.8846,
     "train_samples": 204017,
-    "train_samples_per_second": 145.469,
-    "train_steps_per_second": 2.273
+    "train_samples_per_second": 85.21,
+    "train_steps_per_second": 1.331
 }

trainer_state.json

@@ -1,8 +1,8 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 3.0,
-  "global_step": 9564,
+  "epoch": 2.999745123027154,
+  "global_step": 9561,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -10,19 +10,19 @@
     {
       "epoch": 0.16,
       "learning_rate": 0.0001,
-      "loss": 4.7835,
+      "loss": 4.7889,
       "step": 500
     },
     {
       "epoch": 0.31,
       "learning_rate": 0.0001,
-      "loss": 4.0029,
+      "loss": 4.007,
       "step": 1000
     },
     {
       "epoch": 0.47,
       "learning_rate": 0.0001,
-      "loss": 3.8273,
+      "loss": 3.8226,
       "step": 1500
     },
     {
@@ -34,106 +34,106 @@
     {
       "epoch": 0.78,
       "learning_rate": 0.0001,
-      "loss": 3.6487,
+      "loss": 3.643,
       "step": 2500
     },
     {
       "epoch": 0.94,
       "learning_rate": 0.0001,
-      "loss": 3.5852,
+      "loss": 3.5814,
       "step": 3000
     },
     {
       "epoch": 1.1,
       "learning_rate": 0.0001,
-      "loss": 3.5361,
+      "loss": 3.5365,
       "step": 3500
     },
     {
-      "epoch": 1.25,
+      "epoch": 1.26,
       "learning_rate": 0.0001,
-      "loss": 3.4853,
+      "loss": 3.4833,
       "step": 4000
     },
     {
       "epoch": 1.41,
       "learning_rate": 0.0001,
-      "loss": 3.4501,
+      "loss": 3.4472,
       "step": 4500
     },
     {
       "epoch": 1.57,
       "learning_rate": 0.0001,
-      "loss": 3.4219,
+      "loss": 3.4199,
       "step": 5000
     },
     {
       "epoch": 1.73,
       "learning_rate": 0.0001,
-      "loss": 3.3996,
+      "loss": 3.398,
       "step": 5500
     },
     {
       "epoch": 1.88,
       "learning_rate": 0.0001,
-      "loss": 3.3681,
+      "loss": 3.3639,
       "step": 6000
     },
     {
       "epoch": 2.04,
       "learning_rate": 0.0001,
-      "loss": 3.3408,
+      "loss": 3.3425,
       "step": 6500
     },
     {
       "epoch": 2.2,
       "learning_rate": 0.0001,
-      "loss": 3.3105,
+      "loss": 3.3083,
       "step": 7000
     },
     {
       "epoch": 2.35,
       "learning_rate": 0.0001,
-      "loss": 3.3033,
+      "loss": 3.2997,
       "step": 7500
     },
     {
       "epoch": 2.51,
       "learning_rate": 0.0001,
-      "loss": 3.2689,
+      "loss": 3.2649,
       "step": 8000
     },
     {
       "epoch": 2.67,
       "learning_rate": 0.0001,
-      "loss": 3.2592,
+      "loss": 3.258,
       "step": 8500
     },
     {
       "epoch": 2.82,
       "learning_rate": 0.0001,
-      "loss": 3.2535,
+      "loss": 3.2514,
       "step": 9000
     },
     {
       "epoch": 2.98,
       "learning_rate": 0.0001,
-      "loss": 3.2336,
+      "loss": 3.2301,
       "step": 9500
     },
     {
       "epoch": 3.0,
-      "step": 9564,
-      "total_flos": 1.0309255643489894e+17,
-      "train_loss": 3.5344422363128087,
-      "train_runtime": 4207.4456,
-      "train_samples_per_second": 145.469,
-      "train_steps_per_second": 2.273
+      "step": 9561,
+      "total_flos": 8.150411322232013e+16,
+      "train_loss": 3.5328583012638624,
+      "train_runtime": 7182.8846,
+      "train_samples_per_second": 85.21,
+      "train_steps_per_second": 1.331
     }
   ],
-  "max_steps": 9564,
+  "max_steps": 9561,
   "num_train_epochs": 3,
-  "total_flos": 1.0309255643489894e+17,
+  "total_flos": 8.150411322232013e+16,
   "trial_name": null,
   "trial_params": null
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2f4452815b843d7fbd4d6ccc6494c80d29fe0d2e902d2471d312ebbebceb16fb
-size 3631
+oid sha256:1c4c81cb7753ab52e2ab7a0063a8c2ebcdffab508d0c985f8c7eb5a96bdafd2b
+size 3643