Final One

README.md

@@ -14,14 +14,14 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [csebuetnlp/banglat5](https://huggingface.co/csebuetnlp/banglat5) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- eval_loss: 0.0119
-- eval_Val Accuracy: 0.9843
-- eval_gen_len: 2.2716
-- eval_runtime: 24.164
-- eval_samples_per_second: 92.493
-- eval_steps_per_second: 2.897
-- epoch: 36.0
-- step: 4032
+- eval_loss: 0.3232
+- eval_Val Accuracy: 0.9011
+- eval_gen_len: 2.2671
+- eval_runtime: 18.8367
+- eval_samples_per_second: 118.651
+- eval_steps_per_second: 3.716
+- epoch: 23.0
+- step: 1150
 
 ## Model description
 
@@ -41,7 +41,7 @@ More information needed
 
 The following hyperparameters were used during training:
 - learning_rate: 0.0005
-- train_batch_size: 200
+- train_batch_size: 360
 - eval_batch_size: 32
 - seed: 42
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
@@ -50,7 +50,7 @@ The following hyperparameters were used during training:
 
 ### Framework versions
 
-- Transformers 4.34.1
+- Transformers 4.35.0
 - Pytorch 1.13.1+cu117
 - Datasets 2.14.6
 - Tokenizers 0.14.1

generation_config.json

@@ -2,5 +2,5 @@
   "decoder_start_token_id": 0,
   "eos_token_id": 1,
   "pad_token_id": 0,
-  "transformers_version": "4.34.1"
+  "transformers_version": "4.35.0"
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:dfc97e28e6dfcb9e44059191181add42394313315129852577e687985d57bd84
+oid sha256:c7e683b84cc18b7c8129d72142e58f8824292175fbfa2deb6e27c583638cf2cc
 size 990345064

runs/Nov08_06-49-19_a035082b428d/events.out.tfevents.1699426163.a035082b428d.6281.0

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:15a295881e55ef0193a84e598267582f700379cc6fb4cef9a0a037d9bfe1fc46
-size 12045
+oid sha256:31362eb03a4243d4ac978fcbd1ef361cc23398a1ef0c2c8f1b43f8e2ecd03ad5
+size 13557