Model save

README.md

@@ -1,7 +1,7 @@
 ---
 library_name: transformers
 license: apache-2.0
-base_model: google/flan-t5-base
+base_model: allmalab/gpt2-aze
 tags:
 - generated_from_trainer
 model-index:
@@ -14,10 +14,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # aztelecom-complaint-resolver
 
-This model is a fine-tuned version of [google/flan-t5-base](https://huggingface.co/google/flan-t5-base) on an unknown dataset.
-It achieves the following results on the evaluation set:
-- Loss: 0.6030
-- Generation Quality: 0.0
+This model is a fine-tuned version of [allmalab/gpt2-aze](https://huggingface.co/allmalab/gpt2-aze) on an unknown dataset.
 
 ## Model description
 
@@ -37,11 +34,11 @@ More information needed
 
 The following hyperparameters were used during training:
 - learning_rate: 3e-05
-- train_batch_size: 8
-- eval_batch_size: 8
+- train_batch_size: 4
+- eval_batch_size: 2
 - seed: 42
 - gradient_accumulation_steps: 4
-- total_train_batch_size: 32
+- total_train_batch_size: 16
 - optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
 - lr_scheduler_type: cosine
 - lr_scheduler_warmup_ratio: 0.1
@@ -49,10 +46,6 @@ The following hyperparameters were used during training:
 
 ### Training results
 
-| Training Loss | Epoch  | Step | Validation Loss | Generation Quality |
-|:-------------:|:------:|:----:|:---------------:|:------------------:|
-| 0.8492        | 4.6729 | 500  | 0.7531          | 1.4626             |
-| 0.6632        | 9.3458 | 1000 | 0.6030          | 0.0                |
 
 
 ### Framework versions

generation_config.json

@@ -1,6 +1,6 @@
 {
-  "decoder_start_token_id": 0,
-  "eos_token_id": 1,
-  "pad_token_id": 0,
+  "_from_model_config": true,
+  "bos_token_id": 11,
+  "eos_token_id": 12,
   "transformers_version": "4.49.0"
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cb2357db4d5dbc8b50b73aee33258be5078a83ac5942fbb18c3611064b0ba01e
+oid sha256:7f6a436412a1ae2b1d80130315db6a394d6c73e0daf8a271c69f56ce618b2250
 size 540001920