Model save

README.md

@@ -13,7 +13,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [](https://huggingface.co/) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 6.6845
+- Loss: 6.7011
 
 ## Model description
 
@@ -38,7 +38,7 @@ The following hyperparameters were used during training:
 - seed: 42
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
-- lr_scheduler_warmup_steps: 80
+- lr_scheduler_warmup_steps: 20
 - num_epochs: 30
 - mixed_precision_training: Native AMP
 
@@ -46,8 +46,8 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch   | Step  | Validation Loss |
 |:-------------:|:-------:|:-----:|:---------------:|
-| 6.513         | 11.9617 | 5000  | 6.8757          |
-| 6.057         | 23.9234 | 10000 | 6.6845          |
+| 6.4968        | 11.9617 | 5000  | 6.8573          |
+| 6.0917        | 23.9234 | 10000 | 6.7011          |
 
 
 ### Framework versions

best/config.json

@@ -6,7 +6,7 @@
   "dim_feedforward": 1024,
   "dropout": 0.1,
   "input_dim": 30000,
-  "max_seq_len": 850,
+  "max_seq_len": 512,
   "model_type": "transformer",
   "nhead": 8,
   "num_decoder_layers": 6,

best/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7a3b5e8b8b5618e2453da38fc47c9c32c4aa9d5c037cea4a928d64f5174eb071
-size 252674096
+oid sha256:96ee9b2481257248ae8cd26863390d9f6fdee382cdbaf3d00db433eb2e1db109
+size 251289632

best/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:65d2e3895171c7c111dca323509e114a62b8d79e477d1b1893443b467389d585
+oid sha256:7670b3b289ae6a643928d39e22c9f21dd6d4ca010011bee8559985ee2019845a
 size 5112